Contents lists available at ScienceDirect

Journal of Asian Earth Sciences

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

Full length article

Do Hf isotopes in magmatic zircons represent those of their host rocks?

Di Wanga,b, Xiao-Lei Wanga,⁎, Yue Caib, Steven L. Goldsteinb, Tao Yanga

a State Key Laboratory for Mineral Deposits Research, School of Earth Sciences and Engineering, Nanjing University, Nanjing 210046, Chinab Lamont-Doherty Earth Observatory of Columbia University, Department of Earth and Environmental Sciences, Columbia University, Palisades, NY 10964, USA

A R T I C L E I N F O

Keywords:Hf isotopesDissolution experimentsDisequilibrium meltingDecouplingZircon

A B S T R A C T

Lu–Hf isotopic system in zircon is a powerful and widely used geochemical tracer in studying petrogenesis ofmagmatic rocks and crustal evolution, assuming that zircon Hf isotopes can represent initial Hf isotopes of theirparental whole rock. However, this assumption may not always be valid. Disequilibrium partial melting ofcontinental crust would preferentially melt out non-zircon minerals with high time-integrated Lu/Hf ratios andgenerate partial melts with Hf isotope compositions that are more radiogenic than those of its magma source.Dissolution experiments (with hotplate, bomb and sintering procedures) of zircon-bearing samples demonstratethis disequilibrium effect where partial dissolution yielded variable and more radiogenic Hf isotope composi-tions than fully dissolved samples. A case study from the Neoproterozoic Jiuling batholith in southern Chinashows that about half of the investigated samples show decoupled Hf isotopes between zircons and the bulkrocks. This decoupling could reflect complex and prolonged magmatic processes, such as crustal assimilation,magma mixing, and disequilibrium melting, which are consistent with the wide temperature spectrum from∼630 °C to ∼900 °C by Ti-in-zircon thermometer. We suggest that magmatic zircons may only record the Hfisotopic composition of their surrounding melt during crystallization and it is uncertain whether their Hf isotopiccompositions can represent the primary Hf isotopic compositions of the bulk magmas. In this regard, using zirconHf isotopic compositions to trace crustal evolution may be biased since most of these could be originally fromdisequilibrium partial melts.

1. Introduction

Lu–Hf isotopic system has been widely used as a petrogenetic tracer,yielding information on time-integrated fractionation of Lu and Hfduring processes of melting, crystallization, metasomatism, and as-similation (e.g., Patchett, 1983; Blichert-Toft and Puchtel, 2010;Vervoort, 2015). Lu–Hf isotopic system provides a powerful comple-ment to the widely used Sm–Nd isotopic system, owing to similar butgreater parent-daughter elemental fractionation during magmatic pro-cesses. However, due to the slow diffusion rate of Hf in certain minerals(e.g., zircons) (Cherniak et al., 1997), Hf isotopes of melt would behighly sensitive to the dissolution rate of these minerals, and suchdisequilibrium behavior of Hf presents an important caveat in the ap-plication of Lu–Hf isotopic system (Tang et al., 2014). In particular, theeffect of isotopic disequilibrium between the partial melt and its magmasource may be significant for Hf isotopic compositions of high-viscosityintermediate-felsic magmas because some minerals can crystallize overa long period in the crystal mush and may record transient composi-tions of fractional melt over time rather than the magmatic body or itsmagma source. In recent decades, owing to the rapid development of

laser ablation multi-collector inductively couple plasma mass spectro-meter (LA-MC-ICP-MS) that enabled fast and accurate Hf isotope ana-lyses on small-sized samples, a large number of Hf isotopic studies werecarried out on Hf-enriched minerals, such as zircon, baddeleyite, eu-dialyte, zirconolite and calzirtite (e.g., Wu et al., 2006, 2010; Baueret al., 2017). Zircon is one of the most widely used minerals to inferprimary 176Hf/177Hf compositions of magmas, owing to its high Hfconcentration (commonly 0.5–2.0 wt%; Wu et al., 2006) and extremelylow Lu/Hf ratio (typically ∼0.002; Kinny and Maas, 2003). However,systematic studies are needed to examine the differences in Hf isotopesbetween zircons and their primary magma in order to correctly inter-pret the fast growing dataset of zircon Hf isotopes. Analyzing whole-rock samples is also suitable to acquire Hf isotopes. However, besidesthe difficulty in achieving complete decomposition of zircon-bearingrocks, a few other caveats need to be considered when using zircon orwhole-rock Hf isotopic compositions to infer those of the magma or thesource.

It has been noted that Hf isotopic compositions in zircons frommany magmatic bodies (in particular intermediate-felsic intrusions)show larger ranges of variations than analytical uncertainties, even

https://doi.org/10.1016/j.jseaes.2017.12.025Received 11 December 2017; Received in revised form 12 December 2017; Accepted 13 December 2017

⁎ Corresponding author.E-mail address: wxl@nju.edu.cn (X.-L. Wang).

Journal of Asian Earth Sciences 154 (2018) 202–212

Available online 14 December 20171367-9120/ © 2017 Elsevier Ltd. All rights reserved.

T

within individual hand specimens (e.g., Tang et al., 2014). These var-iations have been attributed to four potential reasons: (1) differentzircons may have crystallized from different batches of melts withdifferent Hf isotopic compositions during the melt accumulation pro-cess of intermediate–felsic magmas (e.g., Kemp et al., 2007; Shaw andFlood, 2009); (2) when partial melting happens at a low temperatureand/or over a short time period, zircons could be inherited from thesource (of the protolith) and they may not be fully dissolved so that theacquired whole-rock Hf isotopes would be biased to infer Hf isotopiccompositions of primary melt (e.g., Mahar et al., 2016); (3) under high-grade metamorphism, transient metamorphic fluids could impart theirHf isotope compositions on all or part of zircons during zircon growthor overgrowth (e.g., Zheng et al., 2005); and (4) during magma mixingor crustal assimilation, zircons may record Hf isotopic compositions ofthe evolving melt while solidified magmas represent a final integratedcomposition (e.g., Boss, 2008; Yang et al., 2015; Couzinié et al., 2016).As Hf isotopic compositions in zircon could be affected by these mag-matic processes, they may not simply be coupled with those of thewhole-rock. Results from Ti-in-zircon thermometer indicate that mag-matic zircons could crystallize over a long period during magmaticevolution (e.g., Harrison et al., 2007). More importantly, high precisionsingle crystal CA-ID-TIMS ages demonstrate very convincingly thecrystallization of magmatic zircon over a protracted time period insingle magmas (e.g., Glazner et al., 2004; Schoene et al., 2012; Barboniet al., 2015; Szymanowski et al., 2017). Even in high-temperature low-viscosity ultramafic-mafic magmas, magmatic zircons may not recordHf isotopic compositions of the primary melt, because they tend tocrystallize later during magma evolution and record progressive con-tamination (Wang et al., 2016).

To accurately interpret zircon Hf isotopic data, it is necessary toevaluate the effects of these processes on the coupling and decouplingof Hf isotopic compositions between zircons and their host rocks. Tothis end, we carried out two studies: (1) dissolution experiments toensure total dissolution of whole-rock samples and confirm the effect ofdisequilibrium melting on Hf isotopic compositions; and (2) a casestudy to compare zircon vs. whole-rock Hf isotopic analyses of theNeoproterozoic Jiuling batholith of southern China.

2. Geological background and sample descriptions

The Jiuling batholith is located in the eastern part of JiangnanOrogen. It is the largest Neoproterozoic granitic intrusion in southernChina (Fig. 1a), with an exposed area of more than 2500 km2 (Wanget al., 2013a and references therein) and zircon U–Pb ages of835–810Ma (e.g., Li et al., 2003; Wang et al., 2013a; Zhao et al., 2013).The Jiuling batholith could be divided into three units based on li-thology and mineral size: Jiuxiantang (JXT), Shihuajian (SHJ) andJiuling (JL) units (Fig. 1b). The JL unit is composed of medium- andcoarse-grained granodiorites (35–55% plagioclase, 25–35% quartz,3–15% K-feldspar, 5–15% biotite, 0–5% cordierite and 0–5% garnet).The SHJ unit is also composed of granodiorites, but, with fine grains.The SHJ granitoids show mineral assemblages of 40–55% plagioclase,25–35% quartz, 5–40% K-feldspar, 5–10% biotite and 0–5% cordierite.In comparison, the JXT unit is composed of tonalites, with mineralassemblages of 40–55% plagioclase, 25–35% quartz, 2–7% K-feldsparand 10–15% biotite. Biotite is the dominant mafic minerals for therocks and no hornblende has been observed in the batholith. Twenty-three granitoids samples were collected from the Jiuling batholith. Thegranitic rocks of the Jiuling batholith are overall homogeneous withSiO2 contents of 65.2–71.2 wt% (Li et al., 2003). Locally we can seesome microgranular enclaves, and they were suggested to representmixing of supracrustal melts (i.e. partial melting of sedimentary rocks)(Zhao et al., 2013) or xenoliths of metasedimentary rocks (Xu et al.,1993). Only four granitic samples (09JL-06-1, 09JL-08-1, 09JL-14-2and 07JL-06-2) were taken from the outcrops where some enclaves canbe observed. Some garnets were observed in samples 11JL-04, 11JL-06-

1, 11JL-16, 11JL-19 and 11JL-21, and some cordierites were observedin samples 09JL-08-1 and 07JL-06-2. The appearance of garnet andcordierite could also indicate the incorporation of supracrustal mate-rials (e.g. sedimentary rocks) (Wang et al., 2013a).

To carry out the decomposition experiments, we selected fivesamples with mafic–felsic lithology based on preliminary data. Onegabbroic sample (11ZJ-10-1) was collected from the Huangshan plutonin northwestern Zhejiang Province (ca. 820Ma; Wang et al., 2012). Onegabbroic sample (08XL-10-1) and one dioritic sample (08XL-06-4) werecollected from the Xialan complex in northern Guangdong Province (ca.195Ma; Yu et al., 2009). One dioritic sample (11JH-01-1) was selectedfrom the Miaohou complex in central Zhejiang Province (ca. 830Ma;Xia et al., 2015). One granitic standard (GSR-1 from the Institute ofGeophysical and Geochemical Exploration, China) was from the Qian-lishan granitic pluton of southern Hunan Province (ca. 150Ma; Li et al.,2004). These samples were selected and used in the decompositionexperiments because the CL images and geochronological data haveshown that they contain very rare old zircon xenocrysts or inheritedzircons which make them suitable for the evaluation of the effect ofdisequilibrium partial melting processes. A summary of the ages, pet-rography and geochemistry for each of the representative samples arelisted in Table 1.

3. Analytical methods

For zircon Hf isotope analysis, zircon grains were separated fromcrushed samples using conventional heavy liquid and magnetic tech-niques, mounted in epoxy resin and polished down to approximatelyhalf-section thickness to expose the grain centers. Isotopic and con-centration analyses were carried out based on cathodoluminescence(CL) images as well as transmitted and reflected light micrographs. CLimages were taken using a Mono CL3+ (Gatan, USA) attached to ascanning electron microscope (Quanta 400 FEG) at the State KeyLaboratory of Continental Dynamics, Northwest University, Xi’an.

3.1. Zircon U–Th–Pb and trace element concentration

Zircon trace element concentrations and U–Th–Pb isotopic compo-sitions were mostly analyzed using an Agilent 7500a ICP-MS attachedto a New Wave 213 nm laser ablation system (later attached to a GeoLasPro 193 nm laser ablation system) at the State Key Laboratory forMineral Deposits Research, Nanjing University (MiDeR-NJU), followingthe procedures of Wang et al. (2007). Few LA-ICP-MS zircon U–Pbdating were also conducted at Shandong Bureau of China MetallurgicalGeology Bureau. To correct U–Pb fractionation, zircon standardGEMOC GJ-1 (207Pb/206Pb age of 608.5 ± 1.5Ma; Jackson et al. 2004)was measured. NIST-612 glass was used for trace element concentrationcalculation with 29Si as the internal standard. Zircon Mud Tank stan-dard (732 ± 5Ma; Black and Gulson, 1978) and basalt glass standardKL2G (Kilauea tholeiite; Jochum et al., 2000) were routinely analyzedto ensure age and trace element concentration accuracy, respectively.All analyses were carried out using a repetition rate of 5 Hz or 7 Hz.According to the size of zircon, the laser ablation beams were 25 µm or35 µm in diameter. U–Pb ages were calculated using the on-line soft-ware package GLITTER (ver. 4.4) (Griffin et al., 2008; www.mq.edu.au/GEMOC). The common Pb correction was carried out through the Excelprogram of CompbCorr#3_15G (Andersen, 2002). U–Th–Pb age calcu-lations and Concordia diagrams were generated using the ISOPLOT/Exprogram (ver. 2.06) of Ludwig (1999).

3.2. Zircon Lu–Hf isotopes

After zircon U–Pb and trace element analyses, in situ zircon Lu–Hfisotopic analyses were obtained on the same spots or in the same do-mains as U–Pb analyses, using a New Wave ArF 193 nm laser ablationsystem attached to a Neptune plus multiple-collector inductively

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

203

coupled plasma mass spectrometer (MC-ICP-MS) at the MiDeR-NJU,following the procedures of Griffin et al. (2006) and Wang et al.(2013b). Typical ablation time is about 30 s for 200 cycles at laser beamdiameters of 35 μm and repetition rates of 5 Hz. In order to monitoranalytical accuracy, reference zircon Mud Tank (176Hf/177Hf =0.282507 ± 6; Woodhead and Hergt, 2005) and zircon standard91,500 (Wiedenbeck et al., 1995) were analyzed routinely. Initial Hfisotope ratios (in terms of εHf(t)) are calculated based on 176Lu decayconstant of 1.867×10−11 per year (Söderlund et al., 2004) and176Hf/177Hf = 0.282785 and 176Lu/177Hf = 0.0336 (Bouvier et al.,2008) as chondritic values. Single-stage model ages (TDM1) and two-stage “crustal” model ages (TDM2) were calculated assuming a present-day depleted mantle (DM) 176Hf/177Hf ratio of 0.28325 (Nowell et al.,1998) and a 176Lu/177Hf ratio of 0.0384 (Griffin et al. 2002); and theTDM2 values were further calculated using measured U–Pb ages and abulk crust 176Lu/177Hf ratio of 0.015 (Griffin et al., 2002).

3.3. Whole-rock Hf isotopes

Whole-rock Hf isotopic compositions were measured by ThermoScientific Neptune (Plus) multiple-collector inductively coupled plasmamass spectrometers (MC-ICP-MS) at the MiDeR-NJU (acid hotplate di-gestion and pressurized bomb digestion methods) and at Lamont-Doherty Earth Observatory (LDEO) (Na2O2 sintering method). Thesample digestion procedures are shown as following.

3.3.1. Hot-plate procedureEight sample materials (100mg/each), combined with 0.1 mL

HClO4 and 2mL HF, were dissolved in capped 60mL Savillex™ Teflonscrew-top vessels. The vessels were heated on hot plate at 120 °C for7 days. After that, the solution was evaporated to near dryness andfumed on hot plate at 200 °C. Sequentially, add 1mL 6M HCl intovessels and dry samples up for twice. Then sample materials in 5mL3M HCl were capped in vessels and put on hot plate at 80 °C for 12 hr.

3.3.2. Bomb procedureEach sample was dissolved in a Teflon capsule with 1.5mL con-

centrated hydrofluoric acid (HF). After evaporating to near dryness, theweighted samples were re-dissolved with 1.5 mL concentrated HF and1.0 mL concentrated HNO3. After that, these sealed Teflon capsules

were put in airtight steel sleeves and heated at ∼190 °C in an oven forvariable time. They were heated on a hotplate to dry up after adding1mL concentrated HNO3 three times. Upon cooling, the capsule wasopened and heated to dry up. 1mL of 6M HCl was subsequently addedto the residue and evaporated, and this procedure was then repeatedtwice. When it had cooled down, the residue was dissolved in 5mL of3M HCl. The capsules were sealed again and placed on a hotplate at∼100 °C overnight (12 hr). The dissolved samples were centrifugallyseparated from the precipitation. The bomb procedure is similar to themethod used by Yang et al. (2010).

3.3.3. Sintering procedure100mg sample was thoroughly mixed with more than 600mg of

granular sodium peroxide (Na2O2), and covered by a thin layer ofgranular Na2O2 in glassy carbon crucibles. The crucibles sequentiallywere heated for 30min at 490 °C in the muffle furnace. Upon cooling toroom temperature, they were covered with watch glass in order toprevent cross contamination. Then water was added dropwise until theexothermal reaction ceased. The suspension solution was transferred tocentrifuge tubes to be centrifuged at high speed for 20min. The su-pernatant was discharged to remove soluble components (mainly Naand Si). Hydrochloric acid was added to the supernatant to test whethersilica gel is evolving, indicated by blurring of the liquid. Then theprecipitate was dissolved by adding first 1 mL of concentrated hydro-chloric acid and transferred to Teflon vessels to dry up. Then dissolvesamples in 5mL 6M HCl–0.06M HF on a hot plate at 100 °C for onehour. After drying down again, take up samples in 5mL 3M HCl andcentrifuge for 10min. Add ∼1mL 1M ascorbic acid–3M HCl mixture,agitate the sample solution until yellow color fade away, and thencentrifuge them again for 20min. The sintering procedure was adaptedfrom Kleinhanns et al. (2002).

We found that with our pressurized vessel, 96 hr is sufficient toachieve full decomposition for granitoids from the Jiuling batholith. Hfpurification procedures are similar to those described by Yang et al.(2010) and Münker et al. (2001), respectively. In order to monitor theaccuracy and precision of our procedures, Certified Reference Materials(CRMs) of mafic to felsic rocks were analyzed together with the sam-ples. Moreover, five representative mafic, intermediate and felsicsamples were also analyzed to test the effect of zircon decomposition onwhole-rock Hf isotopes. These five reference rock materials were

Fig. 1. (a) Simplified geological map of the South China Block (modified after Wang et al., 2013b). (b) Geological map of the Neoproterozoic Jiuling batholith in northwestern JiangxiProvince (modified after Wang et al., 2013a).

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

204

Table1

Petrolog

ical

andge

oche

mical

inform

ationof

represen

tative

samples

from

theNeo

proteroz

oicJiulingba

tholith.

Unit

Sample

Roc

ktype

Grt

Cor

MEs

Age

(Ma)

2σSiO2

(wt%

)Zr

(ppm

)Zircon

εHf(t)

2σZircon

ΔεHf(t)a

Who

le-roc

kεH

f(t)b

2σWho

le-roc

kεH

f(t)c

2σdεHf |(

WR–Z

rn)/WR|d

ΔTTi-in-

zrce

(°C)

Who

le-

rock

εNd(t)

2σGPS

Location

JXT

09JL

-06-1

Tona

lite

✓81

85

63.0

122

4.9

0.8

6.9

2.8

0.2

3.2

0.2

0.74

−3.3

0.1

N28

°46′17

.3″

E11

4°57

′41.9″

JXT

09JL

-08-1

Tona

lite

✓✓

814

667

.711

53.8

1.0

5.8

1.9

0.1

1.04

111

−4.5

0.1

N28

°47′22

.7″

E11

4°57

′02.5″

SHJ

09JL

-13-1

Grano

diorite

825

663

.411

94.6

2.2

8.7

4.7

0.1

0.02

−2.3

0.1

N28

°33′21

.1″

E11

4°51

′32.7″

SHJ

09JL

-14-2

Grano

diorite

✓82

26

68.5

138

2.7

1.8

9.8

0.0

0.1

80.2

−5.4

0.1

N28

°33′36

.7″

E11

4°42

′55.5″

SHJ

09JL

-15-1

Grano

diorite

822

769

.214

43.5

3.7

14.2

4.1

0.1

5.1

0.2

0.15

−4.3

0.1

N28

°33′30

.2″

E11

4°41

′53.4″

SHJ

11JL

-10

Granite

811

1474

.710

73.6

6.0

5.0

3.2

0.2

3.7

0.1

0.14

39−3.8

0.3

N28

°39′46

.1″

E11

4°36

′07.1″

JL07

JL-06-2

Grano

diorite

✓✓

816

565

.618

12.9

1.4

7.3

1.7

0.1

1.6

0.2

0.67

68N

28°36′59

.0″

E11

5°00

′44.7″

JL07

JL-09-3

Grano

diorite

814

569

.294

.83.0

1.3

14.1

3.0

0.1

0.01

111

−3.3

0.1

N28

°38′06

.5″

E11

5°01

′13.6″

JL09

JL-5-1

Grano

diorite

815

566

.413

23.4

0.8

4.4

−1.5

0.1

3.19

184

−3.6

0.1

N28

°43′52

.8″

E11

4°58

′44.6″

JL09

JL-11-1

Grano

diorite

817

469

.411

64.5

1.0

6.4

4.6

0.1

5.0

0.1

0.06

222

−1.7

0.2

N28

°33′23

.8″

E11

4°51

′36.3″

JL11

JL-03-1

Grano

diorite

823

1267

.321

90.4

1.4

7.6

1.0

0.1

0.7

0.3

0.61

−3.2

0.1

N28

°25′49

.4″

E11

4°33

′48.7″

JL11

JL-04

Grano

diorite

✓81

06

65.7

230

−2.0

1.1

5.8

2.0

0.1

1.8

0.1

2.10

−3.6

0.3

N28

°26′37

.1″

E11

4°33

′34.5″

JL11

JL-06-1

Grano

diorite

✓82

210

68.4

208

−1.4

2.0

8.5

0.2

0.1

1.0

0.2

7.13

−3.7

0.1

N28

°29′14

.7″

E11

4°18

′37.4″

JL11

JL-11-2

Grano

diorite

819

1170

.794

.31.7

1.0

4.9

3.1

0.2

3.8

0.1

0.44

−3.7

0.3

N28

°45′06

.9″

E11

4°39

′37.5″

JL11

JL-12

Grano

diorite

820

1165

.518

42.7

0.7

3.5

3.5

0.1

3.8

0.2

0.23

−2.2

0.1

N28

°51′41

.8″

E11

4°45

′52.5″

JL11

JL-14-2

Granite

816

773

.270

.05.9

1.0

5.0

7.0

0.2

0.16

131

−3.1

0.3

N28

°52′33

.1″

E11

4°51

′40.6″

JL11

JL-16

Grano

diorite

✓81

66

65.4

247

0.6

1.1

9.1

2.7

0.1

2.9

0.2

0.67

214

−2.5

0.1

N28

°49′54

.4″

E11

4°55

′09.3″

JL11

JL-17

Grano

diorite

821

1270

.814

2−

0.5

0.9

3.8

3.3

0.1

3.0

0.2

1.15

−2.8

0.1

N28

°52′34

.9″

E11

5°00

′47.2″

JL11

JL-19

Grano

diorite

✓81

66

67.5

220

3.2

2.2

9.2

1.4

0.1

2.1

0.1

1.25

107

−2.8

0.1

N28

°59′45

.9″

E11

5°07

′23.1″

JL11

JL-20

Grano

diorite

818

1066

.815

30.4

3.8

18.1

3.2

0.1

0.88

−3.4

0.3

N28

°57′26

.5″

E11

5°10

′45.2″

JL11

JL-21

Grano

diorite

✓✓

822

1166

.523

50.7

1.2

7.7

2.0

0.1

2.7

0.2

0.65

−3.0

0.1

N28

°50′53

.2″

E11

5°40

′15.8″

JL11

JL-22A

Grano

diorite

820

667

.920

63.0

1.1

6.3

2.3

0.1

2.7

0.2

0.28

123

−2.2

0.1

N28

°50′11

.7″

E11

5°42

′09.8″

JL11

JL-23

Grano

diorite

827

569

.616

13.5

1.0

8.4

3.1

0.1

3.6

0.1

0.14

127

−2.6

0.1

N28

°46′24

.2″

E11

5°45

′30.0″

Note:

Grt

=Garne

t,Cor

=Cordierite,

MEs

=microgran

ular

enclav

es;

Zircon

ΔεHf(t)a=

Zircon

εHf(t)max

−Zircon

εHf(t)min;

b,crepresen

tin

thebo

mban

dsinteringproc

edures,r

espe

ctively;

dεHf |(

WR–Z

rn)/WR|d

represen

tstheeq

uation

:|(w

hole-roc

kεH

f(t)-zirco

nεH

f(t))/who

le-roc

kεH

f(t)|;

ΔTTi-in-zrce=

(TTi-in-zrc) m

ax−

(TTi-in-zrc) m

in.

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

205

obtained from the United States Geological Survey (USGS), i.e. W-2A(diabase), BCR-2 (basalt), AGV-2 (andesite) and GSP-1& -2 (grano-diorite); one from Russian Academy of Sciences (RAS), i.e. SG-3(granite); one from Geological Survey of Japan (GSJ), i.e. JR-2 (rhyo-lite); and one in-house standard K1919 (i.e. BHVO-1; basalt) from Ha-waiian basalt.

In order to evaluate the reproducibility and accuracy of the in-strument, a JMC-475 Hf standard solution of 100 ppb was used. Eachrun consisted of 4–5 blocks of 10 cycles per block and 3 blocks of 20cycles per block, respectively, at MiDeR-NJU and LDEO. Isobaric in-terferences (i.e. 176Yb and 176Lu on 176Hf) were corrected based on thestable ratios: 176Yb/173Yb = 0.79323 and 176Lu/175Lu = 0.026528.After subtraction of the isobaric interferences, the Hf measurements areinternally corrected for mass bias using the stable ratio of 179Hf/177Hf(0.7325), suggested by Rehkämper et al. (2001) and based on the ex-ponential law (Russell et al., 1978). Analysis of the Hf standard JMC-475 during the course of this study yielded 176Hf/177Hf =0.282173 ± 14 (n= 12; MiDeR-NJU) and 176Hf/177Hf =0.282144 ± 9 (n= 24; LDEO), broadly identical to the recommendedvalues (0.282163 ± 9 by Blichert-Toft et al., 1997 and0.282161 ± 16 by Goolaerts et al., 2004) within uncertainties. Thereported 176Hf/177Hf ratios were further adjusted relative to the JMC-475 standard of 0.282160 (Nowell et al., 1998).

Whole-rock Nd isotopic compositions were analyzed using theNeptune (Plus) MC-ICP-MS at the MiDeR-NJU, with detailed analyticalprocedures are similar to that of Wang et al. (2017).

4. Results

4.1. Hf isotopes of representative mafic–felsic samples in dissolutionexperiments

Hf isotopic data of Certified Reference Materials (CRMs) are listed inTable S1 in the Electronic Supplementary Appendix A. All analyzedunknowns are normalized to JMC 475 value of 0.282160 (Vervoort andBlichert-Toft, 1999). Mafic-felsic standard analyses yield reliable meanHf isotopic compositions consistent with reported values (Fig. 2). Theseexperiments confirm the precision and accuracy of our analytical pro-cedures from two different labs, without the complication of zircondissolution.

Hf isotopic data of representative mafic–felsic samples are listed inTable S2 in the Electronic Supplementary Appendix A. Whole-rock Hfisotopes analyzed by bomb dissolution method show the maximumdifferences in 176Hf/177Hf ratios of 0.000074, 0.000028, 0.000019,0.000094 and 0.000015 for samples 11ZJ-10-1, 08XL-10-1, 08XL-06-4,11JH-01-1 and GSR-1, which correspond to the maximum εHf(0) (εHf isthe deviation in parts per 10,000 from a “chondritic uniform reservoir”(CHUR) value) differences of 2.6, 1.0, 0.7, 3.3 and 0.5 epsilon units,respectively (Fig. 3). Based on the observed systematic differences in Hfisotope ratios among the solutes from different dissolution experiments,we divided the bomb dissolution processes into two stages. The 1ststage represents the first 48 hr dissolution, during which only part ofthe zircons from the whole rocks were dissolved, resulting in elevatedHf isotopic compositions in the solute than the whole rock (based onsintered dissolution and longer period of bomb dissolutions).176Hf/177Hf ratios of the solutes gradually decrease as the length of thebomb dissolution time increased, up to 48 hr. This is consistent withintensified decomposition of refractory zircons that retains un-radiogenic Hf. The 2nd stage represents decomposition time after 48 hr,during which Hf isotopic compositions of the solutes become re-producible and consistent with those from sintered dissolution method(Fig. 3). After 48 hr of bomb dissolution, all samples yielded consistent176Hf/177Hf ratios that fall within analytical uncertainty, with narrowεHf(0) differences of 1.8, 1.0, 0.6, 0.4 and 0.3 epsilon units, respectively.This suggests that bomb dissolution with decomposition time longerthan 48 hr and sintering procedures can both fully dissolve zircon-

bearing rock samples. Moreover, Hf isotopic compositions of these twodissolution methods show excellent linear correlation with a slope of 1,and a correlation coefficient of 0.91 (Table S3 in the Electronic Sup-plementary Appendix A; Fig. 4), which also support the reliability ofsintering and bomb dissolution procedures (> 48 hr).

4.2. Zircon and whole-rock Hf isotopes of the Neoproterozoic Jiulingbatholith

Zircon grains in samples from the Jiuling batholith commonly showaspect ratios of 1:1–4:1. They exhibit euhedral to subhedral mor-phology and clear oscillatory zoning, indicating a magmatic origin(Hoskin et al., 2003). They contain Th and U concentrations rangingfrom 23 to 1309 ppm and from 50 to 1535 ppm, respectively, with Th/U ratios mostly from 0.20 to 2.64 (Table S4 in the Electronic Supple-mentary Appendix A). All the magmatic zircons yield 206Pb/238U dateswithin the range of 796 ± 18Ma to 839 ± 22Ma (Table S4), which isconcordant for each sample and consistent with the published ages ofthe Jiuling batholith (e.g., Wang et al., 2013a; Zhao et al., 2013). TheHf isotope ratios and the U–Pb dates measured from spot analyses wereused to calculate their initial Hf isotopic ratios and εHf(t) values. Wefound that magmatic zircons from the Neoproterozoic Jiuling batholithshow a large range of initial 176Hf/177Hf ratios from 0.281955 to0.282558 and corresponding εHf(t) in the range from −10.7 to +8.8(Table S5 in the Electronic Supplementary Appendix A). In contrast,most of the whole-rock analyses of the Jiuling batholith yield positiveεHf(t) values from +0.0 to +7.0, except for one sample (09JL-5-1) withεHf(t) value of −1.5. It is worth noting that Ti-in-zircon thermometry ofmagmatic zircons from the Jiuling batholith records a large tempera-ture range of 633–905 °C. This large temperature range, together withthe large range of zircon Hf isotopic ratios, may reflect the long periodof magmatic evolution during which the zircons crystallized.

4.3. Whole-rock Nd isotopes from the Neoproterozoic Jiuling batholith

All rock samples exhibit negative εNd(t) values (εNd is the deviationin parts per 10,000 from a CHUR value) from −5.4 to −1.7.Specifically, within the SHJ unit, sample 09JL-14-2 shows the minimumvalue of −5.4; in the JXT unit, sample 09JL-8-1 shows the low value of–4.5; in the JL unit, samples 09JL-5-1, 11JL-04, 11JL-06-1 and 11JL-11-2 are characterized with more negative εNd(t) values (–3.6 to –3.7) thanthose of the others (–1.7 to –3.3). The six samples have more negativeεNd(t) than the other samples from the same pluton, which indicatemore incorporation of ancient crustal materials. Below we will discussthe data further.

5. Discussion

5.1. Can disequilibrium melting affect zircon Hf isotopes?

It has been suggested that disequilibrium melting with respect to Hfisotopes can be promoted by the presence of zircons in magma sources(Tang et al., 2014). To understand the dissolution behavior of zircongrains during partial melting, we carried out dissolution experiments ofzircon-bearing rocks which could be considered as an analogue forcrustal melting regarding zircons. Obviously, dissolution with acid isdifferent from natural melting process. Here it is important to show thatpartial dissolution of certain minerals would contribute the variation ofHf isotopes in the resultant solution/melt. It is also reasonable tocompare the two processes because zircons are refractory during bothmelting and dissolution processes while mafic minerals that contain Luand Hf concentrations (such as biotite and hornblende) tend to dissolveor breakdown (melt) prior to zircons.

The dissolution experiments show that the hotplate method gen-erally yields significantly higher 176Hf/177Hf ratios than the other twomethods, especially for intermediate and felsic rocks (Fig. 3). For

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

206

example, the maximum difference in 176Hf/177Hf ratios between hot-plate and sintering methods are 0.000415 and 0.000075 for dioritesample 11JH-01-1 and granite sample GSR-1, respectively, with cor-responding difference in εHf(0) (hereafter referred to as ΔεHf(0)) of 14.7and 2.7 epsilon units, respectively. Note that the two samples are ex-pected to contain abundant zircons owing to their high SiO2 contentsand zirconium concentrations (Table S2). In the 1st stage of bombmethod, 176Hf/177Hf ratios gradually decrease with increasing bombdissolution time, which likely reflects increasing extents of decom-position of zircons. During the 2nd stage of bomb dissolution, Hf iso-topic compositions keep stable and consistent with those of sinteringmethod, which indicates that all zircons have been dissolved in thisstage. It is noted that complete dissolution leads to lower Hf isotopesthan partial dissolution as zircons have lower Lu/Hf ratios.

Partial dissolution of zircon-bearing rocks could occur duringcrustal melting, as what inherited zircons showed in many graniticrocks (e.g., Hanchar et al., 2003). Similarly, intensified decompositionof zircons during crustal melting would also generate lower 176Hf/177Hf

ratios in the melt and as a result, different degrees of partial meltingcould generate melts with different Hf isotope ratios. This effect couldbe particularly important during transport and emplacement of low-temperature, high viscosity granitic magmas. During the initial stagesof melting of older rocks, when zircons are not completely dissolved,the partial melt would have relatively high Lu/Hf ratios and radiogenic176Hf/177Hf ratios and high εHf(t) values; during later stages of melting,more zircons are dissolved into the melt, which leads to decreasing Lu/Hf ratios and unradiogenic 176Hf/177Hf ratios and low εHf(t) values inthe partial melt (Fig. 5). The fractional partial melts should follow asimilar temporal trend as the solutes from the 1st stage of the bombdissolution (Fig. 3). New crystallized zircons from different fractionalpartial melts can only equilibrated with their surrounding melts. Duringmagma advection, zircons from different batches may show distinct Hfisotopic compositions, which may generate heterogeneous Hf isotopecompositions within individual zircons from one hand specimen.

Fig. 2. Comparison of corrected 176Hf/177Hf ratios for Certified Reference Materials (CRMs) with reported reference values. Data sources: BCR-2 from Bizzarro et al. (2003), Connellyet al. (2006), Hanyu et al. (2005), Le Fèvre and Pin (2001), McCoy-West et al (2010), Vervoort et al. (2004), Weis et al. (2007) and Yang et al. (2010); BHVO-1 from Connelly et al.(2006), Pourmand and Dauphas (2010), Reagan et al. (2010), Ulfbeck et al. (2003) and Weis et al. (2007); AGV-2 from Aciego et al. (2009), Khanna et al. (2014), Pourmand and Dauphas(2010), Weis et al. (2007) and Yang et al. (2010); GSP-1 from Mahlen et al. (2008); GSP-2 from Weis et al. (2007); JR-2 from Hanyu et al. (2005).

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

207

5.2. Coupling and decoupling of zircon and whole-rock Hf isotopes: Possiblemechanisms

Magmatic zircon Hf isotopes have been widely used as the initial Hfisotopes of melts to trace magma sources, and related crustal evolution

and reworking history (e.g., Bauer et al., 2017). However, as discussedabove, disequilibrium fractional melting may complicate the inter-pretation of these results. Although Patchett (1983) first proposed thatconcordant to slightly discordant zircons appear to be suitable and re-liable carriers of initial Hf isotopes, he also pointed out that combinedzircon and whole rock Hf isotopes are recommended until the processesthat control the addition of radiogenic Hf to zircons are understood.Within our studied samples from the Jiuling batholith, all magmaticzircons are concordant in U–Th–Pb isotopes, without evident Pb loss,precluding the influence of post-magmatic Hf loss or addition as Hfdiffuses more slowly than Pb in zircons (Cherniak and Watson, 2001).Therefore, the observed variations of Hf isotopes in magmatic zirconscould be used to investigate the potential factors that contribute to thecoupling and decoupling of Hf isotopes between zircons and hostmagma.

The coupling and decoupling of Hf isotopes can be identified insamples from the Jiuling batholith. Only ∼45% (ten out of twenty-three) of studied samples exhibit coupled Hf isotopes between zirconsand their host rocks (Fig. 6a). Interestingly, the decoupling for the othersamples has different characteristics. In some cases, zircons showhigher 176Hf/177Hf than that of the whole-rock (Fig. 6a), such as sam-ples 09JL-06-1 and 09JL-08-1 from the JXT unit, sample 09JL-14-2from the SHJ unit and samples 07JL-06-2, 09JL-5-1 and 11JL-19 fromthe JL unit; conversely, whole-rock Hf isotopes are higher than zirconsin other samples, such as the samples from the JL unit (e.g., sample11JL-04, 11JL-06-1 and 11JL-17). However, it is noted that sampleswith large ranges of Ti concentrations and calculated Ti-in-zircon

Fig. 3. Corrected 176Hf/177Hf ratios for variable time in bomb method and ratios in hotplate and sintering methods. In the legend, HB represents the same decomposition effect at acertain time within the first 12 hr using bomb method as that of hotplate method. Red arrows represent the 176Hf/177Hf variation trends for all samples.

Fig. 4. Comparison of whole-rock εHf(t) values between the bomb and sintering methods.The data shows a good correlation between them.

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

208

thermometer (TzirconTi ) results generally show no relationship with the

decoupling between zircons and whole-rocks (Fig. 6b). This is con-sistent with our hypothesis that the decoupling could be generatedduring long-term magmatic evolution as recorded by Tzircon

Ti . In addition,a slight positive correlation is visible between whole-rock Zr con-centration and the deviation of Hf isotopic compositions in zircons fromcorresponding host rocks (hereafter denoted as dεHf|(WR–Zrn)/WR| inFig. 6c). The large temperature ranges and high Zr concentrations in thewhole rock indicate a long crystallization period for zircons, both ofwhich could facilitate complex magmatic processes and diverse zirconformations (i.e. antecrysts, autocrysts, xenocrysts and inherited zircons;Miller et al., 2007). Additionally, there is no obvious relationship be-tween zircon ΔεHf(t) (where ΔεHf(t) = εHf(t)max− εHf(t)min) anddεHf|(WR–Zrn)/WR| when all the samples are considered (Fig. 6d), in-dicating that large zircon ΔεHf(t) variation could result from complex

processes, including inheritance of older zircons, crystallization se-quences, crustal assimilation, and magma mixing. For example, oncesignificant assimilation occurred, a larger difference between zirconsand host rock is expected, resulting in larger zircon ΔεHf(t) anddεHf|(WR–Zrn)/WR| values. Given this assumption, the lack of assimilationand/or mixing could be a possible proxy for Hf isotopic coupling be-tween zircons and host rocks. Additional data is needed to test thisproxy and perhaps better proxies could be identified to screen for de-coupled zircon-whole rock Hf isotopes.

As discussed earlier, disequilibrium fractional melting can lead de-coupling in Hf isotopes between zircons and their host rock.Additionally, zircon crystallization over a long period of time could alsogenerate decoupling. Based on Tzircon

Ti results, zircons could formthroughout the history of magmatic evolution during crustal melting(Miller et al., 2007). Under microscope, zircons in samples 11JL-14 and

Fig. 5. Schematic diagrams illustrating 176Hf/177Hf variations due to different degrees of zircon decomposition in magma (i.e. disequilibrium melting). In the early stage (a), partialmelting of sources leads to dissolution of few zircons; the generated melts are characterized with significantly higher Lu/Hf and 176Hf/177Hf ratios than those of sources in average. In thefollowing stage (b), more zircon decomposition results in slightly higher Lu/Hf ratios and Hf isotopes in melts than those in sources. In these two stages, new generated zircons havehigher εHf(t) than those of sources (i.e. whole rock). In the late stage (c), few residual zircons in sources contribute more Lu concentration to melts with Hf isotopes approximately equal tothose of sources. In this stage, newly-formed zircons have broadly equal εHf(t) with those of sources (i.e. whole rock).

Fig. 6. (a) Comparison of zircon εHf(t) values and those of whole rocks; (b) dεHf|(WR–Zrn)/WR| vs. TΔ zirconTi (°C); (c) dεHf|(WR–Zrn)/WR| vs. Zr concentrations; (d) dεHf|(WR–Zrn)/WR| vs. zircon

ΔεHf(t). Here, dεHf|(WR–Zrn)/WR| = |(whole-rock εHf(t)–zircon εHf(t))/whole-rock εHf(t)|, which can reflect the deviation degree of zircon Hf isotopes from whole rock ones. Note onesample (09JL-14-2) was excluded in Fig. 5b–d due to the abnormal large dεHf|(WR–Zrn)/WR| value. Assimilated samples are assimilation-affected samples and indicated by more negativeεNd(t) values; assimilated/mixing samples are indicated by the appearance of xenoliths or enclaves in field and garnet and/or cordierite under microscope in our samples. The errors ofwhole-rock εHf(t) are too small to be visible.

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

209

09JL-5-1 are both enwrapped in early-forming minerals (such as eu-hedral–subhedral biotite and K-feldspar) and late-forming minerals(such as small xenomorphic quartz), which possibly implies that zirconscould form at different magmatic stages. Though it may be difficult todistinguish zircon antecrysts (zircons that crystallized earlier in themagmatic history) from autocrysts (zircons that crystallized late in themagmatic history) based on ages alone (as the ages may overlap withinerrors), they may record different Hf isotopes, which may not representthat of the magmatic body as a whole.

Decoupling of Hf isotopic composition can also result from mixingor assimilation of exotic crust-derived components. Xenoliths and mi-crogranular enclaves with elliptoid and irregular shapes appear at ornear some of our sample locations (samples 09JL-06-1, 09JL-08-1,09JL-14-2 and 07JL-06-2). These enclaves likely formed during magmamixing of supracrustal melts or assimilation of the NeoproterozoicShuangqiaoshan Group metasedimentary country rocks (e.g., Wanget al., 2013a; Zhao et al., 2013). Notably, these four samples alsocontain magmatic zircons with higher εHf(t) values than those of thewhole-rock. Wang et al. (2013a) proposed that the appearance of garnetand cordierite in the Jiuling batholith could be the product of in-corporation of supracrustal materials. Seven samples (e.g., sample11JL-04, 11JL-06-1, 11JL-16, 11JL-19 & 11JL-21) show garnet and/orcordierite, which may indicate assimilation or mixing with supracrustalmaterials. They also show decoupled zircon vs. whole-rock Hf isotopes.Moreover, the assimilated samples, such as sample 09JL-5-1 and 11JL-04 indicated by Nd isotopes, also show relative low zircon εHf(t),compared with the whole rocks. Thus, assimilation and magma mixingplay an important role in controlling the decoupling behaviors. In thissituation, zircons antecrysts could retain Hf isotopes of the magma priorto assimilation, while zircon autocrysts could record information ofassimilation or mixing with variable εHf(t) values, which would furtherincrease the difference between whole rock and zircon Hf isotopic ra-tios.

There is another possibility that Hf isotopic decoupling could also beinherited from magma source. Almost all magmas are isotopicallyheterogeneous on different scales from zoned crystals to map-scale

zoning (e.g., Beard et al., 2005; Gao et al., 2016). This source hetero-geneity could show the time evolved εHf arrays of the magmatic rimsoverlap closely with the εHf range displayed by the inherited cores (witholder ages) at the crystallization age as indicated by Villaros et al.(2012). Only in some samples (i.e. sample 11JL-11-2 & 11JL-20), εHf(t)of some magmatic zircons are similar to those of inherited zircon cores(with older ages) calculated at the crystallization age of the batholith,which could imply a genetic relationship between them and indicateheterogeneity in the magma source. Though these samples also exhibitdecoupled zircon and whole rock Hf isotopes, the heterogeneity ofsources could not account for most of decoupled samples. Thus, het-erogeneous melting could not play a significant role in the decouplingof Hf isotopes between zircon and whole rocks.

In these cases, zircon Hf isotopic compositions may not simply re-flect whole-rock Hf isotopes. Furthermore, Hf isotopic compositions ofzircons or whole rock composition of one individual hand samplecannot be simply used to represent the Hf isotopic compositions ofmagma sources.

5.3. Implications for magmatic evolution

The coupling and decoupling of Hf isotopes between zircons andhost rocks can also shed some light on our understanding on magmaticprocesses. It has been suggested that a large pluton can accumulate overmillions of years (e.g., Coleman et al., 2004). Previous geochemical andgeochronological studies also suggested multiple magma batches forthe formation of the huge Jiuling batholith (e.g., Li et al., 2003; Zhaoet al., 2013) although the duration of emplacement is not exactlyknown because of the large uncertainties of the reported crystallizationages. The observed variations in Hf isotopes and the decoupling of Hfisotopes between zircon and whole-rock of this study strengthen a longand complex magma accumulation history.

The combined dissolution experiments and case study on the Jiulingbatholith can shed light on the evolution of granitic body. In summary,the magmatic evolution history, as it pertains to Hf isotopes, should beevaluated in multiple stages. At first, the magmatic processes need to beconstrained based on the melting behavior of magma source. Differentmagma batches can be generated by different degrees of partial meltingof different sources. Magma sources are commonly heterogeneous sothat the resulted melts would inherit such heterogeneity. Fractionalmelts extracted from the sources would have different Hf isotopes dueto variable melting degree and different residual mineral assemblagesin sources (Fig. 7). Both physical mixing and conductive heating couldtrigger some extents of partial melting of country rocks surrounding themagma chamber and addition of country rock Hf to the magma. Fol-lowing magma extraction, additional heterogeneity could be introducedto the magma through magma mixing and the assimilation of countryrocks, and the heterogeneity could be recorded by new crystallizedzircons from the evolving magma (e.g., Paterson et al., 2008; Wotzlawet al., 2015) (Fig. 7). Thus, neither zircon nor whole-rock Hf isotopesalone could capture the complexity of the magmatic processes, espe-cially for intermediate to felsic plutonic rocks. However, coupled orsimilar initial Hf isotopic compositions in zircons and whole-rocks maymake a strong case for zircon-free source and/or equilibrium melting.Combination of the zircon and whole-rock Hf isotopes is therefore va-luable for reconstructing melting and melts transport processes.

Additionally, as detrital zircons are mostly eroded from inter-mediate to felsic magmatic rocks that may show large Hf isotopic var-iations, the inherent complexities of their host rocks need to be con-sidered while using Hf isotopic compositions in detrital zircons toreconstruct their crustal source. One possible way to eliminate thiscomplexity is to combine oxygen isotope analysis with Hf isotopeanalyses in detrital zircons as shown by Hawkesworth and Kemp(2006), where elevated oxygen isotope reflects assimilation of crustalrock.

Fig. 7. A schematic diagram illustrating possible mechanism that leads to decoupled Hfisotopes between zircon and whole rocks. The different degree of partial melting (i.e.disequilibrium melting) causes the Hf diversity in melts (with different Hf isotopes de-noted as εHf1, εHf2 and εHf3 at different magmatic stages). The magma experiences long-lived and complex crystallization progresses, involving assimilation and mixing (with Hfcharacters of εHf4), which would further lead more heterogeneity in Hf isotopic com-positions. The generated melts aggregate and ascend to form a magma chamber withεHf5. In such stages, zircon antecrysts, autocrysts, xenocrysts and inherited zircons pre-serve various Hf isotopes, distinct from those of whole rock. All these factors could ac-count for the Hf isotopic decoupling (See text for detailed explanation).

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

210

6. Conclusions

The decomposition experiments in this study provide evidence thatabnormally high Hf isotopes and large isotopic variation (from severalto a dozen epsilon units) could result from disequilibrium melting ofzircon-bearing crustal rocks. Hf isotopic ratios of the zircons could notalways couple with those of their host rocks, especially when their hostrocks are granitoids. Decoupling between whole rock and zircon Hfisotopes can result from complex magmatic progresses, such as assim-ilation, mixing and disequilibrium melting, during a prolonged zircon-crystallization history. Furthermore, neither zircon nor whole rock Hfisotopes alone may faithfully record the 176Hf/177Hf ratios of magma,except when εHf(t) of the zircons are similar to that of the whole rock.The lack of assimilation and/or mixing may be useful proxies to identifycoupled zircon and whole-rock Hf isotopes. When Hf isotopes of thezircons and the whole rock agree with each other, it would yield reli-able Hf isotopes of the magma. These considerations imply that therange of Hf isotopic variations recorded by detrital zircons may notsimply reflect different proportions of contributions from differentsources because zircons derived from intermediate-felsic plutonicbodies may already have a large Hf isotopic diversity.

Acknowledgments

This work was substantially supported by the State Key R&D Projectof China (2016YFC0600203), the National Natural Science Foundationof China (41472049), and the China Scholarship Council (CSC, to D.Wang). The manuscript benefited a lot from previous anonymous re-viewers and the editor Prof. M.F. Zhou. We thank L. Bolge, A.J. Lin, X.Yan, and Z.Y. Zhu for whole-rock Hf isotope analyses.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in theonline version, at https://doi.org/10.1016/j.jseaes.2017.12.025.

References

Aciego, S.M., Bourdon, B., Lupker, M., Rickli, J., 2009. A new procedure for separatingand measuring radiogenic isotopes (U, Th, Pa, Ra, Sr, Nd, Hf) in ice cores. Chem.Geol. 266, 194–204.

Andersen, T., 2002. Correction of common lead in U-Pb analyses that do not report 204Pb.Chem. Geol. 192, 59–79.

Barboni, M., Annen, C., Schoene, B., 2015. Evaluating the construction and evolution ofupper crustal magma reservoirs with coupled U/Pb zircon geochronology andthermal modeling: a case study from the Mt. Capanne pluton (Elba, Italy). EarthPlanet. Sci. Lett. 432, 436–448.

Bauer, A.M., Fisher, C.M., Vervoort, J.D., Bowring, S.A., 2017. Coupled zircon Lu–Hf andU-Pb isotopic analyses of the oldest terrestrial crust, the> 4.03 Ga Acasta GneissComplex. Earth Planet. Sci. Lett. 458, 37–48.

Beard, J.S., Ragland, P.C., Crawford, M.L., 2005. Reactive bulk assimilation: a model forcrust-mantle mixing in silicic magmas. Geology 33, 681–684.

Bizzarro, M., Baker, J.A., Ulfbeck, D., 2003. A new digestion and chemical separationtechnique for rapid and highly reproducible determination of Lu/Hf and Hf Isotoperatios in geological materials by MC-ICP-MS. Geostand. Newsl. 27, 133–145.

Black, L.P., Gulson, B.L., 1978. The age of the Mud Tank carbonatite, strangways range,Northern Territory. BMR J. Aust. Geol. Geophys. 3, 227–232.

Blichert-Toft, J., Puchtel, I.S., 2010. Depleted mantle sources through time: evidence fromLu–Hf and Sm–Nd isotope systematics of Archean komatiites. Earth Planet. Sci. Lett.297, 598–606.

Blichert-Toft, J., Chauvel, C., Albarède, F., 1997. Separation of Hf and Lu for high pre-cision isotope analysis of rock samples by magnetic sector-multiple collector ICP–MS.Contrib. Mineral. Petrol. 127, 248–260.

Boss, A.P., 2008. Mixing in the solar nebula: implications for isotopic heterogeneity andlarge-scale transport of refractory grains. Earth Planet. Sci. Lett. 268, 102–109.

Bouvier, A., Vervoort, J.D., Patchett, P.J., 2008. The Lu–Hf and Sm–Nd isotopic compo-sition of CHUR: constraints from unequilibrated chondrites and implications for thebulk composition of terrestrial planets. Earth Planet. Sci. Lett. 273, 48–57.

Cherniak, D.J., Hanchar, J.M., Watson, E.B., 1997. Diffusion of tetravalent cations inzircon. Contrib. Mineral. Petrol. 127, 383–390.

Cherniak, D.J., Watson, E.B., 2001. Pb diffusion in zircon. Chem. Geol. 172, 5–24.Coleman, D.S., Gray, W., Glazner, A.F., 2004. Rethinking the emplacement and evolution

of zoned plutons: geochronologic evidence for incremental assembly of the TuolumneIntrusive Suite, California. Geology 32, 433–436.

Connelly, J.N., Ulfbeck, D.G., Thrane, K., Bizzarro, M., Housh, T., 2006. A method forpurifying Lu and Hf for analyses by MC-ICP-MS using TODGA resin. Chem. Geol. 233,126–136.

Couzinié, S., Laurent, O., Moyen, J.F., Zeh, A., Bouilhol, P., Villaros, A., 2016. Post-col-lisional magmatism: crustal growth not identified by zircon Hf–O isotopes. EarthPlanet. Sci. Lett. 456, 182–195.

Gao, P., Zheng, Y.F., Zhao, Z.F., 2016. Distinction between S-type and peraluminous I-type granites: zircon versus whole-rock geochemistry. Lithos 258, 77–91.

Glazner, A.F., Bartley, J.M., Coleman, D.S., Gray, W., Taylor, R.Z., 2004. Are plutonsassembled over millions of years by amalgamation from small magma chambers?GSA Today 14 (14), 4–11.

Goolaerts, A., Mattielli, N., Jong, J.D., Weis, D., Scoates, J.S., 2004. Hf and Lu isotopicreference values for the zircon standard 91500 by MC–ICP–MS. Chem. Geol.206, 1–9.

Griffin, W.L., Belousova, E.A., Walters, S.G., O’Reilly, S.Y., 2006. Archean and Proterozoiccrustal evolution in the eastern succession of the Mt Isa district Australia: U-Pb andHf-isotope studies of detrital zircons. Aust. J. Earth Sci. 53, 125–149.

Griffin, W.L., Powell, W.J., Pearson, N.J., O’Reilly, S.Y., 2008. GLITTER: data reductionsoftware for laser ablation ICP–MS. In: Sylvester, P (Eds), Laser Ablation–ICP–MS inthe Earth Sciences, Mineralogical Association of Canada Short Course Series 40, pp.204–207.

Griffin, W.L., Wang, X., Jackson, S.E., Pearson, S.E., O’Reilly, S.Y., Xu, X.S., Zhou, X.M.,2002. Zircon chemistry and magma genesis, SE China: in situ analysis of Hf isotopes,Tonglu and Pingtan igneous complexes. Lithos 61, 237–269.

Hanchar, J.M., Watson, E.B., 2003. Zircon saturation thermometry. Rev. Mineral.Geochem. 53, 89–112.

Hanyu, T., Nakai, S.I., Tatsuta, R., 2005. Hafnium isotope ratios of nine GSJ referencesamples. Geochem. J. 39, 83–90.

Harrison, T.M., Watson, E.B., Aikman, A.B., 2007. Temperature spectra of zircon crys-tallization in plutonic rocks. Geology 35, 635–638.

Hawkesworth, C.J., Kemp, A.I.S., 2006. Using hafnium and oxygen isotopes in zircons tounravel the record of crustal evolution. Chem. Geol. 226, 144–162.

Hoskin, P.W., Schaltegger, U., 2003. The composition of zircon and igneous and meta-morphic petrogenesis. Rev. Mineral. Geochem. 53, 27–62.

Jochum, K.P., Dingwell, D.B., Rocholl, A., Stoll, B., Hofmann, A.W., Becker, S., Besmehn,A., Bessette, D., Dietze, H.J., Dulski, P., Erzinger, J., Hellebrand, E., Hoppe, P., Horn,I., Janssens, K., Jenner, G.A., Klein, M., McDonough, W.F., Maetz, M., Mezger, K.,Müker, C., Nikogosian, I.K., Pickhardt, C., Raczek, I., Rhede, D., Seufert, H.M.,Simakin, S.G., Sobolev, A.V., Spettel, B., Straub, S., Vincze, L., Wallianos, A.,Weckwerth, G., Weyer, S., Wolf, D., Zimmer, M., 2000. The preparation and pre-liminary characterisation of eight geological MPI-DING reference glasses for in-situmicroanalysis. Geostand. Newslett. 24, 87–133.

Kemp, A.I.S., Hawkesworth, C.J., Foster, G.L., Paterson, B.A., Woodhead, J.D., Hergt,J.M., Gary, C.M., Whitehouse, M.J., 2007. Magmatic and crustal differentiation his-tory of granitic rocks from Hf-O isotopes in zircon. Science 315, 980–983.

Khanna, T.C., Bizimis, M., Yogodzinski, G.M., Mallick, S., 2014. Hafnium–neodymiumisotope systematics of the 2.7 Ga Gadwal greenstone terrane, Eastern Dharwar craton,India: implications for the evolution of the Archean depleted mantle. Geochim.Cosmochim. Acta 127, 10–24.

Kinny, P.D., Maas, R., 2003. Lu–Hf and Sm–Nd isotope systems in zircon. Rev. Mineral.Geochem. 53, 327–341.

Kleinhanns, I.C., Kreissig, K., Kamber, B.S., Meisel, T., Nagler, T.F., Kramers, J.D., 2002.Combined chemical separation of Lu, Hf, Sm, Nd, and REEs from a single rock digest:precise and accurate isotope determinations of Lu–Hf and Sm–Nd using multi-collector-ICP-MS. Anal. Chem. 74, 67–73.

Le Fèvre, B., Pin, C., 2001. An extraction chromatography method for Hf separation priorto isotopic analysis using multiple collection ICP-mass spectrometry. Anal. Chem. 73,2453–2460.

Li, X.H., Liu, D., Sun, M., Li, W.X., Liang, X.R., Liu, Y., 2004. Precise Sm–Nd and U-Pbisotopic dating of the supergiant Shizhuyuan polymetallic deposit and its hostgranite, SE China. Geol. Mag. 141, 225–231.

Li, X.H., Li, Z.X., Ge, W.C., Zhou, H.W., Li, W.X., Liu, Y., Wingate, M.T.D., 2003.Neoproterozoic granitoids in South China: crustal melting above a mantle plume at ca825 Ma? Precambr. Res. 122, 45–83.

Ludwig, K.R., 1999. User’s Manual for Isoplot/Ex Version 2.06: a GeochronologicalToolkit for Microsoft Excel. (Special Publication). Geochronology Center, Berkeley,pp. 1–49.

Mahar, M.A., Mahéo, G., Goodell, P.C., Pavlis, T.L., 2016. Timing and origin of migmatiticgneisses in south Karakoram: insights from U-Pb, Hf and O isotopic record of zircons.J. Asian Earth Sci. 120, 1–16.

Mahlen, N.J., Beard, B.L., Johnson, C.M., Lapen, T.J., 2008. An investigation of dissolu-tion methods for Lu–Hf and Sm–Nd isotope studies in zircon- and garnet-bearingwhole-rock samples. Geochem. Geophy. Geosyst. 9, Q01002. http://dx.doi.org/10.1029/2007GC001605.

Miller, J.S., Matzel, J.E., Miller, C.F., Burgess, S.D., Miller, R.B., 2007. Zircon growth andrecycling during the assembly of large, composite arc plutons. J. Volcanol. Geotherm.Res. 167, 282–299.

Münker, C., Weyer, S., Scherer, E., Mezger, K., 2001. Separation of high field strengthelements (Nb, Ta, Zr, Hf) and Lu from rock samples for MC-ICPMS measurements.Geochem. Geophys. Geosyst 2 2001GC000183.

Nowell, G.M., Kempton, P.D., Noble, S.R., Fitton, J.G., Saunders, A.D., Mahoney, J.J.,Taylor, R.N., 1998. High precision Hf isotope measurements of MORB and OIB bythermal ionization mass spectrometry: insights into the depleted mantle. Chem. Geol.149, 211–233.

Patchett, P.J., 1983. Importance of the Lu–Hf isotopic system in studies of planetarychronology and chemical evolution. Geochim. Cosmochim. Acta 47, 81–91.

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

211

Paterson, S.R., Pignotta, G.S., Farris, D., Memeti, V., Miller, R.B., Vernon, R.H., Žák, J.,2008. Is stoping a volumetrically significant pluton emplacement process ?Discussion. Geol. Soc. Am. Bull. 120, 1075–1079.

Pourmand, A., Dauphas, N., 2010. Distribution coefficients of 60 elements on TODGAresin: application to Ca, Lu, Hf, U and Th isotope geochemistry. Talanta 81, 741–753.

Reagan, M.K., Ishizuka, O., Stern, R.J., Kelley, K.A., Ohara, Y., Blichert-Toft, J., Bloomer,S.H., Cash, J., Fryer, P., Hanan, B.B., Hickey-Vargas, R.H., Ishii, T., Kimura, J.I.,Peate, D.W., Rowe, M.C., Woods, M., 2010. Fore-arc basalts and subduction initiationin the Izu-Bonin-Mariana system. Geochem. Geophy. Geosyst 11, Q03X12. http://dx.doi.org/10.1029/2009GC002871.

Rehkämper, M., Schönbächler, M., Stirling, C.H., 2001. Multiple collector ICP–MS: in-troduction, measurement techniques and analytical capabilities. Geostand. Newslett.25, 23–40.

Russell, W.A., Papanastassiou, D.A., Tombrello, T.A., 1978. Ca isotope fractionation onthe Earth and other solar system materials. Geochim. Cosmochim. Acta 42,1075–1090.

Schoene, B., Schaltegger, U., Brack, P., Latkoczy, C., Stracke, A., Günther, D., 2012. Ratesof magma differentiation and emplacement in a ballooning pluton recorded by U-PbTIMS-TEA, Adamello batholith, Italy. Earth Planet. Sci. Lett. 355, 162–173.

Shaw, S.E., Flood, R.H., 2009. Zircon Hf isotopic evidence for mixing of crustal and silicicmantle-derived magmas in a zoned granite pluton, eastern Australia. J. Petrol. 50,147–168.

Söderlund, U., Patchett, P.J., Vervoort, J.D., Isachsen, C.E., 2004. The 176Lu decay con-stant determined by Lu–Hf and U-Pb isotope systematics of Precambrian mafic in-trusions. Earth Planet. Sci. Lett. 219, 311–324.

Szymanowski, D., Wotzlaw, J.F., Ellis, B.S., Bachmann, O., Guillong, M., von Quadt, A.,2017. Protracted near-solidus storage and pre-eruptive rejuvenation of large magmareservoirs. Nature Geosci. 10 (10), 777–782.

Tang, M., Wang, X.L., Shu, X.J., Wang, D., Yang, T., Gopon, P., 2014. Hafnium isotopicheterogeneity in zircons from granitic rocks: geochemical evaluation and modeling of“zircon effect” in crustal anatexis. Earth Planet. Sci. Lett. 389, 188–199.

Ulfbeck, D., Baker, J., Waight, T., Krogstad, E., 2003. Rapid sample digestion by fusionand chemical separation of Hf for isotopic analysis by MC-ICPMS. Talanta 59,365–373.

Vervoort, J.D., 2015. Lu–Hf Dating: The Lu–Hf Isotope System. In: Rink, W.J., Thompson,J. (Eds.), Encyclopedia of Scientific Dating Methods. Springer, Netherlands, pp.379–390.

Vervoort, J.D., Blichert-Toft, J., 1999. Evolution of the depleted mantle: Hf isotope evi-dence from juvenile rocks through time. Geochim. Cosmochim. Acta 63, 533–556.

Vervoort, J.D., Patchett, P.J., Söderlund, U., Baker, M., 2004. Isotopic composition of Yband the determination of Lu concentrations and Lu/Hf ratios by isotope dilution usingMC-ICPMS. Geochem. Geophy. Geosyst. 5, Q11002. http://dx.doi.org/10.1029/2004GC000721.

Villaros, A., Buick, I.S., Stevens, G., 2012. Isotopic variations in S-type granites: an in-heritance from a heterogeneous source? Contrib. Mineral. Petrol. 163, 243–257.

Wang, D., Wang, X.L., Cai, Y., Chen, X., Zhang, F.R., Zhang, F.F., 2017. HeterogeneousConservation of Zircon Xenocrystsin Late Jurassic Granitic Intrusions within theNeoproterozoic JiulingBatholith, South China: A Magma Chamber Growth Model inDeep Crustal Hot Zones. J. Petrol. http://dx.doi.org/10.1093/petrology/egx074.

Wang, D., Wang, X.L., Zhou, J.C., Shu, X.J., 2013a. Unraveling the Precambrian crustalevolution by Neoproterozoic conglomerates, Jiangnan orogen: U-Pb and Hf isotopesof detrital zircons. Precambr. Res. 233, 233–236.

Wang, M., Nebel, O., Wang, C.Y., 2016. The Flaw in the Crustal ‘Zircon Archive’: Mixed HfIsotope Signatures Record Progressive Contamination of Late-stage Liquid in Mafic-

Ultramafic Layered Intrusions. J. Petrol. 57, 27–52.Wang, X.L., Zhou, J.C., Griffin, W.L., Wang, R.C., Qiu, J.S., O’Reilly, S.Y., Xu, X.S., Liu,

X.M., Zhang, G.L., 2007. Detrital zircon geochronology of Precambrian basementsequences in the Jiangnan orogen: dating the assembly of the Yangtze Cathaysiablocks. Precambr. Res. 159, 117–131.

Wang, X.L., Zhou, J.C., Wan, Y.S., Kitajima, K., Wang, D., Bonamici, C., Qiu, J.S., Sun, T.,2013b. Magmatic evolution and crustal recycling for Neoproterozoic strongly per-aluminous granitoids from southern China: Hf and O isotopes in zircon. Earth Planet.Sci. Lett. 366, 71–82.

Wang, X.L., Shu, X.J., Xing, G.F., Xie, S.W., Zhang, C.H., Xia, H., 2012. LA-ICP-MS zirconU-Pb ages of the Shijiao-Huangshan intrusive rocks in Zhuji area, Zhejiang Province:implications for the petrogenesis of the ultramafic orbicular rocks. Geol. Bull. China31, 75–81 (in Chinese with English Abstract).

Weis, D., Kieffer, B., Hanano, D., Silva, I.N., Barling, J., Pretorius, W., 2007. Hf isotopecompositions of U.S. Geological Survey reference materials. Geochem. Geophy.Geosyst 8, Q06006. http://dx.doi.org/10.1029/2006GC001473.

Wiedenbeck, M., Alle, P., Corfu, F., Griffin, W.L., Meier, M., Oberli, F., von Quadt, A.,Roddick, J.C., Spiegel, W., 1995. Three natural zircon standards for U-Th–Pb, Lu–Hf,trace element and REE analyses. Geostand. Newslett. 19, 1–23.

Woodhead, J.D., Hergt, J.M., 2005. A preliminary appraisal of seven natural zircon re-ference materials for in situ Hf isotope determination. Geostand. Geoanal. Res. 29,183–195.

Wotzlaw, J.F., Bindeman, I.N., Stern, R.A., D’Abzac, F.X., Schaltegger, U., 2015. Rapidheterogeneous assembly of multiple magma reservoirs prior to Yellowstone super-eruptions. Sci. Rep. 5, 14026. http://dx.doi.org/10.1038/srep14026.

Wu, F.Y., Yang, Y.H., Mitchell, R.H., Bellatreccia, F., Li, Q.L., Zhao, Z.F., 2010. In situ U-Pb and Nd–Hf–(Sr) isotopic investigations of zirconolite and calzirtite. Chem. Geol.277, 178–195.

Wu, F.Y., Yang, Y.H., Xie, L.W., Yang, J.H., Xu, P., 2006. Hf isotopic compositions of thestandard zircons and baddeleyites used in U-Pb geochronology. Chem. Geol. 234,105–126.

Xia, Y., Xu, X.S., Zhao, G.C., Liu, L., 2015. Neoproterozoic active continental margin ofthe Cathaysia block: evidence from geochronology, geochemistry, and Nd–Hf iso-topes of igneous complexes. Precambr. Res. 269, 195–216.

Xu, X.S., Zhou, X.M., Wang, D.Z., 1993. Advances of research on enclaves in inter-mediate–acidic magmatic rocks. Adv. Earth Sci. 4, 008 (in Chinese with EnglishAbstract).

Yang, H., Ge, W.C., Zhao, G.C., Dong, Y., Xu, W.L., Ji, Z., Yu, J.J., 2015. Late Triassicintrusive complex in the Jidong region, Jiamusi-Khanka Block, NE China:Geochemistry, zircon U-Pb ages, Lu–Hf isotopes, and implications for magma min-gling and mixing. Lithos 224, 143–159.

Yang, Y.H., Zhang, H.F., Chu, Z.Y., Xie, L.W., Wu, F.Y., 2010. Combined chemical se-paration of Lu, Hf, Rb, Sr, Sm and Nd from a single rock digest and precise andaccurate isotope determinations of Lu–Hf, Rb–Sr and Sm–Nd isotope systems usingMulti-Collector ICP-MS and TIMS. Int. J. Mass Spectrom. 290, 120–126.

Yu, X., Di, Y., Wu, G., Zhang, D., Zheng, Y., Dai, Y., 2009. The Early Jurassic magmatismin northern Guangdong Province, southeastern China: constraints from SHRIMPzircon U-Pb dating of Xialan complex. Sci. China, Ser. D Earth Sci. 52, 471–483.

Zhao, J.H., Zhou, M.F., Zheng, J.P., 2013. Constraints from zircon U-Pb ages, O and Hfisotopic compositions on the origin of Neoproterozoic peraluminous granitoids fromthe Jiangnan Fold Belt, South China. Contrib. Mineral. Petrol. 166, 1505–1519.

Zheng, Y.F., Wu, Y.B., Zhao, Z.F., Zhang, S.B., Xu, P., Wu, F.Y., 2005. Metamorphic effecton zircon Lu–Hf and U-Pb isotope systems in ultrahigh-pressure eclogite-facies me-tagranite and metabasite. Earth Planet. Sci. Lett. 240, 378–400.

D. Wang et al. Journal of Asian Earth Sciences 154 (2018) 202–212

212