Distinguishing silicate and carbonatite mantle metasomatism by using lithium and its isotopes

Chemical Geology 381 (2014) 67–77

Contents lists available at ScienceDirect

Chemical Geology journal homepage: www.elsevier.com/locate/chemgeo

Distinguishing silicate and carbonatite mantle metasomatism by using lithium and its isotopes Ben-Xun Su a,b,⁎, Hong-Fu Zhang a, Etienne Deloule b, Nathalie Vigier b, Yan Hu c, Yan-Jie Tang a, Yan Xiao a, Patrick Asamoah Sakyi d a

State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, P.O. Box 9825, Beijing 100029, China Centre National de la Recherche Scientifique, CRPG, BP20, 54501 Vandoeuvre-Les-Nancy Cedex, France Isotope Laboratory, Department of Earth and Space Sciences, University of Washington, 38 Seattle, 4000 15th Avenue NE, Seattle, WA 98195, USA d Department of Earth Science, University of Ghana, P.O. Box LG 58, Legon, Accra, Ghana b c

a r t i c l e

i n f o

Article history: Received 8 February 2014 Received in revised form 30 April 2014 Accepted 2 May 2014 Available online 22 May 2014 Editor: K. Mezger Keywords: Li isotopes Mantle xenolith Carbonatite metasomatism Silicate metasomatism Lithospheric mantle

a b s t r a c t To investigate the effects of silicate and carbonatite metasomatism on mantle heterogeneity, we report lithium (Li) concentrations and isotopic compositions for olivine (Ol), orthopyroxene (Opx) and clinopyroxene (Cpx) from two suites of mantle xenoliths (Hannuoba, the North China Craton, and Haoti, the Western Qinling Orogen). The Hannuoba xenoliths range from lherzolite to pyroxenite and were affected by silicate metasomatism, whereas the Haoti xenoliths vary from harzburgite to wehrlite and were affected by carbonatite metasomatism. Lithium concentrations and isotopic compositions display a dichotomy between Hannuoba and Haoti xenoliths, and the overall variation exceeds what was previously reported. The minerals from Haoti xenoliths are more enriched in Li (Ol: 1.23–13.2 ppm; Opx: 3.00–82.8 ppm; Cpx: 1.39–112 ppm) than those from Hannuoba samples (Ol: 1.34–5.52 ppm; Opx: 0.23–16.1 ppm; Cpx: 1.18–79.8 ppm). Lithium isotopic compositions of these samples are highly variable in both suites of samples. δ7Li ranges from + 3.0‰ to + 41.9‰ in Ol, from −21.0‰ to +20.2‰ in Opx and from −17.4‰ to +18.9‰ in Cpx for Hannuoba samples. Haoti minerals display a similar degree of variation with δ7Li ranging from −29.1‰ to +19.9‰ in Ol, −16.9‰ to +18.0‰ in Opx and −45.1‰ to +19.6‰ in Cpx. On average, Li isotopic compositions of minerals from Hannuoba xenoliths follow the sequence of δ7LiOl N δ7LiOpx N δ7LiCpx, whereas those from Haoti xenoliths are characterized by the opposite sequence of δ7LiCpx N δ7LiOpx N δ7LiOl; in particular there is considerable difference in δ7Li values of Ol. The Li elemental and isotopic data suggest that mantle metasomatism by distinct agents is an important process for generating the large heterogeneity of Li abundances and isotopic distribution in the lithospheric mantle. The distinct geochemical characteristics of Li isotopes in silicate and carbonatite metasomatism are closely related to the preferential incorporation of Li into minerals from distinct melts. These findings further demonstrate that the Li isotopic systematics may in turn help to discriminate between silicate and carbonatite metasomatism. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Metasomatism is an important and prevalent mechanism controlling the physical and chemical properties of the mantle through infiltration and percolation of silicate or carbonatite melts and aqueous fluids within the upper mantle (Roden and Murthy, 1985; Dautria et al., 1992). Different consequences can be expected, due to the significant disparity in the physical and chemical properties between silicate and carbonatite metasomatic agents. For example, carbonatite melts have much lower viscosity and density, and greater tendency toward wetting

⁎ Corresponding author at: State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, P.O. Box 9825, Beijing 100029, China. Tel.: +86 10 82998514; fax: +86 10 62010846. E-mail address: [email protected] (B.-X. Su).

http://dx.doi.org/10.1016/j.chemgeo.2014.05.016 0009-2541/© 2014 Elsevier B.V. All rights reserved.

grain boundaries than the silicate melts (Genge et al., 1995; Dobson et al., 1996; Gasparik and Litvin, 2002), hence carbonatite metasomatism can elevate the electrical conductivity of the mantle by 2–3 orders of magnitude compared to silicate metasomatism (Gaillard et al., 2008). Considering their significant roles in the evolutionary history of the mantle and the genesis of basaltic and carbonatitic magmas, it is therefore important to establish geochemical proxies that will aid in deciphering the metasomatic history of mantle samples (Dautria et al., 1992; Laurora et al., 2001; Ying et al., 2004; Halama et al., 2009; Zhang et al., 2009). Silicate metasomatism transforms lherzolite to websterite and orthopyroxenite through olivine (Ol)-consumption and orthopyroxene (Opx)-formation, whereas carbonatite metasomatism would consume Opx to form clinopyroxene (Cpx) and produce a rock series from harzburgite and/or lherzolite to Cpx-rich lherzolite and wehrlite (e.g., Yaxley et al., 1991; Ionov et al., 1996; Laurora et al., 2001). Silicon-

68

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

rich glass inclusion/film is usually present in silicate metasomatized mantle xenoliths, whereas carbonate and CO2-rich fluid inclusions are common phases in carbonatite metasomatized samples (Yaxley et al., 1991; Frezzotti et al., 2002; Su et al., 2012a). Some geochemical indicators such as Ca/Al, La/Yb, Zr/Hf, Nb/Ta and Ti/Eu ratios as well as trace element patterns have also been developed to distinguish between silicate and carbonatite metasomatism (e.g., Yaxley et al., 1991; Green et al., 1992; Veksler et al., 1998; Gorring and Kay, 2000). However, these indicators in many cases can be explained by both processes (e.g., Klemme et al., 1995; Sweeney et al., 1995; Laurora et al., 2001), hence additional indicators are needed to specifically discern between silicate metasomatism and carbonatite metasomatism. Lithium and its isotopes may be such a tracer because of their unique distribution between mantle minerals. Compared to pristine peridotites (Li = 1–1.8 ppm in Ol and 0.50–1.3 ppm in pyroxenes; Seitz et al., 2004; Woodland et al., 2004; Jeffcoate et al., 2007), minerals in metasomatized peridotites have much higher Li concentrations and define larger ranges (Li = 0.6–5 ppm in Ol and 0.6–9.5 ppm in pyroxenes; Woodland et al., 2004; Rudnick and Ionov, 2007; Tang et al., 2010; Zhang et al., 2010). In particular, it has been demonstrated that Li is preferentially incorporated into Ol relative to Cpx and Opx during carbonatite metasomatism, whereas the opposite is true in silicate metasomatism (Seitz and Woodland, 2000; Woodland et al., 2004; Zhang et al., 2010; Tang et al., 2011). Lithium isotopes also display a dichotomy between equilibrated peridotites and metasomatized peridotites. Although a normal mantle is estimated to have a δ7Li of + 1‰ to + 6‰, as sampled by fresh mid-ocean ridge basalts (Jeffcoate et al., 2007; Tomascak et al., 2008), peridotite xenoliths worldwide display a much larger range with Ol varying from − 3‰ to + 15‰ (average value: + 4.7‰), Opx from −26‰ to + 13‰ (average value: +1.1‰) and Cpx from −20‰ to + 13‰ (average value: − 1.1‰) (Su et al., 2012b and references therein). The large Li isotopic variations in these peridotite xenoliths are attributed to metasomatic processes (e.g., Nishio et al., 2004; Wagner and Deloule, 2007; Aulbach and Rudnick, 2009; Mallmann et al., 2009; Tang et al., 2012). Thus, combined studies of Li concentration and isotopic compositions in mantle-derived xenoliths may help to discriminate between silicate and carbonatite metasomatic agents. The actual mechanism of Li transfer from melt to mantle mineral remains unclear, although the importance of diffusion has been addressed in some studies (Rudnick and Ionov, 2007; Wunder et al., 2007). Also, it is unknown how Li isotopic variations depend on the nature of metasomatic agents because this has yet to be investigated and is in part due to the lack of Li isotopic data for carbonatite metasomatized samples. Here, we report in situ Li concentration and isotopic data for two suites of mantle-derived xenoliths from Hannuoba, the North China Craton, and Haoti, the Western Qinling Orogen. These two suites of samples have experienced typical silicate and carbonatite metasomatism, respectively, and thus are ideal for exploring the potential of using Li abundances and isotopes to distinguish between silicate and carbonatite metasomatism.

2. Samples 2.1. Silicate-metasomatized xenoliths from Hannuoba The Hannuoba basalt field is situated at the northern margin of the North China Craton and has been dated at 22 Ma to 14 Ma (Liu et al., 1990). It is well known for its abundant occurrence of diverse mantle xenoliths, varying in composition from peridotite (dunite, harzburgite, lherzolite and wehrlite) to garnet- and/or spinel-bearing pyroxenite (websterite, orthopyroxenite and clinopyroxenite) (Fig. 1a; Chen et al., 2001; Liu et al., 2005; Zhang et al., 2009; Zheng et al., 2009). These xenoliths have been investigated extensively and were interpreted as products of variable degrees of interaction between peridotites and silicate

melts (not host basaltic melts) at a depth of 45–65 km with temperature ranging from 800 °C to 1100 °C. The evidence is as follows: 1) Composite xenoliths (e.g., peridotite with pyroxenite veins) display gradual modal variation in mineralogical composition and reaction textures. Orthopyroxenite, a Si-rich rock type, is abundant (Fig. 1a; Liu et al., 2003, 2005; Zhang et al., 2009). 2) The formation of Opx and the zoning texture in Ol indicate mineral transformations during interaction process, whereas the absence of hydrous minerals and/or secondary Cpx in peridotite and pyroxenite supports silicate melts to be the metasomatic agent (Choi et al., 2008). 3) Systematic trace element and Sr–Nd–Pb–Hf–Os–Li isotopic compositions of whole rocks and mineral separates indicate metasomatic overprinting (Song and Frey, 1989; Liu et al., 2005; Tang et al., 2007; Choi et al., 2008; Zhang et al., 2009; Tang et al., 2012). 4) Si-rich glass melt inclusions are abundantly present in Ol and Cpx in the Hannuoba peridotites (Liu et al., 2003; Du and Fan, 2011). 5) Zircon U–Pb dating and sulfide Re–Os dating of the Hannuoba xenoliths show that the inferred metasomatic agents correspond to regional geological events such as Triassic collision, Late Jurassic– Early Cretaceous post-collisional extension, and Tertiary magmatism (e.g., Liu et al., 2004; Zheng et al., 2009). The xenoliths in this study include two lherzolites, ten websterites, three orthopyroxenites and three clinopyroxenites, covering all types of mantle-derived lithologies in Hannuoba. Their petrological and mineralogical characteristics are similar to those described in previous studies (e.g., Song and Frey, 1989; Chen et al., 2001; Zhang et al., 2013). 2.2. Carbonatite-metasomatized xenoliths from Haoti The Qinling–Dabie Orogenic Belt is tectonically sutured between the North China Craton and Yangtze Craton. It was formed during the closure of the Paleo-Tethyan Ocean and the collision between the North China and Yangtze Cratons from Paleozoic to Mesozoic. Cenozoic kamafugite and carbonatite association dated at 23 Ma to 7 Ma is distributed in the Western Qinling (Yu et al., 2005). Mantle xenoliths collected in the Haoti cinder cones are mainly spinel and garnet facies lherzolites (Fig. 1b; Su et al., 2010a, 2011, 2012a,b,c). Recent studies suggest that the lithospheric mantle beneath the Western Qinling is compositionally stratified with a gradual decrease in fertility with depth, a feature that probably resulted from varying degrees of carbonatite metasomatism (Su et al., 2010a, 2011). Carbonatite metasomatic signatures recorded in these xenoliths are listed as follows: 1) Carbonate veins and discrete grains are commonly observed (Fig. 1b; Su et al., 2009, 2010a, 2012a). 2) Clinopyroxene has high CaO and Na2O contents, has high light rare earth elements (LREE) relative to heavy rare earth elements (HREE), has Ba, Th, U, Pb and Sr enrichments and negative Ti, Hf and Y anomalies, and is in particular characterized by high Ti/Eu ratio (Su et al., 2010a,b, 2011). Additionally, the spongy textures are commonly present in the Cpx of the Haoti xenoliths and have been interpreted to result from decompressional melting (Su et al., 2011). 3) Whole-rock compositions of the Haoti peridotites display LREEenrichment, positive Sr and Ba anomalies, carbonatite-like trace element patterns, and Sr–Nd–Pb isotopic mixing trend between DM (depleted mantle) and EMII (enriched mantle II) end members, which are consistent with the geochemical features resulting from carbonatite metasomatism (Su et al., 2010a, 2012a). 4) Previous Li isotopic studies indicate that the constituent minerals of the Haoti peridotites have extremely high Li contents and distinct δ7Li values, which might be linked to carbonatite metasomatism (Su et al., 2012b). The lherzolite xenoliths investigated here include two garnet-facies, six spinel-facies and one garnet-spinel coexisting samples, which were

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

69

Fig. 1. Petrographic features of metasomatized xenoliths. a) Silicate-metasomatized xenoliths from the Hannuoba area, North China Craton have a series from lherzolite, through olivine websterite, to orthopyroxenite; b) carbonatite-metasomatized xenoliths from the Western Qinling show spongy texture in clinopyroxene, breakdown feature in orthopyroxene and calcite occurrence. Cpx, clinopyroxene; Ol, olivine; Opx, orthopyroxene; Sp, spinel; Ca, calcite.

derived from a depth of 60 km to 120 km with a temperature range between 900 °C and 1200 °C. Detailed petrological and mineralogical descriptions can be found in Su et al. (2009, 2010a,b, 2011, 2012a,b,c). 3. Analytical methods Major element compositions of minerals were determined by wavelength dispersive spectrometry using a JEOL JXA8100 electron probe microanalyzer (EPMA) at the Institute of Geology and Geophysics, Chinese Academy of Sciences. The EPMA analysis was operated at an accelerating voltage of 15 kV and 10 nA beam current, 5 μm beam spot and 10–30 s counting time on peak. Natural (jadeite [NaAlSiO6] for Na, Al and Si, rhodonite [MnSiO3] for Mn, sanidine [KAlSi3O8] for K, garnet [Fe3Al2Si3O12] for Fe, Cr-diopside [(Mg, Cr)CaSi2O6] for Ca, olivine [(Mg, Fe)2SiO4] for Mg) and synthetic (rutile for Ti, 99.7% Cr2O3 for Cr, Ni2Si for Ni) minerals were used for standard calibration. A program based on the ZAF procedure was used for matrix corrections. Typical analytical uncertainty for all of the elements analyzed is better than 1.5%. Lithium contents and isotopic compositions were measured on goldcoated thin-sections using a Cameca IMS-1270 ion microprobe at CRPG, Nancy, France, following established methods (Decitre et al., 2002). The primary beam was 16O− (15–30 nA, 30 μm diameter). Positive secondary ions with excess kinetic energies of 0 ± 20 eV were detected in pulse counting mode. A 180-s presputtering without raster was applied before analysis. The mass spectrometer was operated at a mass resolving power of 1500 (M/ΔM). The primary beam position in the contrast aperture, the magnetic field and the energy offset were automatically centered before each measurement. Thirty cycles were measured with counting times of 12, 4 and 4 s for 6Li, background at the 6.5 mass, and 7Li, respectively. The counting rate for 7Li ranged from 30,000 to

100,000 cps, depending on the Li content of the sample and the primary beam intensity. Lithium isotopic compositions are reported as δ7Li (=[(7Li / 6Li)sample / (7Li/6Li)L-SVEC − 1] × 1000) relative to the standard NIST SRM 8545 (L-SVEC with 7Li / 6Li = 12.0192; Flesch et al., 1973). The instrumental mass fractionation is expressed in δ7Li units: Δi = δ7LiSIMS − δ7LiTIMS. The value of Δi may change between different sessions, owing to variations in the ion probe set-up and to electron multiplier aging (Deloule et al., 1992). The analyses on standard minerals in the study yield homogeneous Li isotopic compositions (Fig. 2) with δ7Li of CpxBZ226 = − 4.6 ± 0.7‰; OlBZ29 = + 4.4 ± 1.3‰; and OpxBZ226 = −4.2 ± 0.5‰, which are consistent with the previously reported values (Zhang et al., 2010; Su et al., 2012b) and the recommended values (− 4.1‰, + 4.4‰ and − 4.2‰ corresponding to δ6Li values of + 4.1‰, − 4.6‰ and + 4.2‰, respectively; Decitre et al., 2002) within analytical error (Table 1). The external 2σ errors of the isotope compositions for both the standards and the samples were less than 2‰. The largest grains of the samples are selected for analyses since they likely record more signals of metasomatism, and the specific

Fig. 2. Standard Li isotopic variation throughout the analyses with 2σ error bars.

70

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

Table 1 Results of standard analyses: δ7Li (‰) values measured by SIMS compared to compositions measured by thermal ionization mass spectrometry (TIMS). Standard minerals

δ7Li (‰) SIMS values

CpxBZ226 OlBZ29 OpxBZ226

−4.1 ± 0.7a +4.6 ± 0.4a −4.2 ± 1.0a

TIMS values (±1‰) −4.7 ± 0.8b +4.4 ± 1.0b −4.2 ± 1.2b

−4.6 ± 0.7c +4.4 ± 1.3c −4.2 ± 0.5c

−4.1 +4.4 −4.2

Note: Each value represents the average of several determinations. TIMS values are from Decitre et al. (2002). All results are normalized to the NIST L-SVEC Li2CO3. a Reported by Zhang et al. (2010). b Reported by Su et al. (2012a). c This study.

size of the selected grain is variable in different samples. Considering the position effect, we chose the mineral grains located in the center of the rounded thin-sections for core-to-rim traverse analyses. The distance between spots depends on the size of the grain. Within analytical

uncertainty, Li contents measured by Cameca IMS-1270 are identical to those obtained from LA-ICP-MS analyses (Su et al., 2010a,b, 2012b) and by chemical separation and MC-ICP-MS analysis results (Tang et al., 2011). As shown in Table 1S, the Mg# of individual Ol, Opx and Cpx grains is homogeneous from core to rim but Li contents and isotopes are significantly heterogeneous. Thus, the large intra- and intermineral Li isotopic variations do not result from matrix effects (Bell et al., 2009). 4. Results 4.1. Hannuoba xenoliths Minerals in the Hannuoba xenoliths show wide Li elemental and isotopic variations and compositional zonation. Lithium concentrations of Ol range from 1.34 ppm to 5.52 ppm with grain averages ranging from 2.15 ppm to 4.95 ppm, whereas δ7Li varies from + 3.0‰ to + 41.9‰ with average values ranging from +6.7‰ to +33.4‰ for the individual

Fig. 3. Histogram of Li concentration and Li isotopic ratios of Ol, Opx and Cpx in the mantle-derived xenoliths from the Hannuoba and Haoti localities. Lithium data from worldwide localities are from Zack et al. (2003), Nishio et al. (2004), Seitz et al. (2004), Magna et al. (2006), Jeffcoate et al. (2007), Rudnick and Ionov (2007), Halama et al. (2007, 2008, 2009), Marks et al. (2007), Tang et al. (2007, 2011), Wagner and Deloule (2007), Ionov and Seitz (2008), Aulbach and Rudnick (2009) and Zhang et al. (2010).

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

71

samples (Table 1S; Fig. 3). They overlap the results of Ol from spinel peridotites determined on MC-ICP-MS and SIMS (1.05 ppm to 1.80 ppm and +1.2‰ to +14.6‰) (Tang et al., 2007) but show wider variations. The cores of most Ol grains display higher Li concentrations and δ7Li values than their rims, while some grains exhibit opposite variations or patterns between Li concentration and Li isotopic composition from core to rim (Fig. 1S). The Opx grains show larger variations in Li concentrations (0.23 ppm to 8.93 ppm; some N10 ppm) and δ7Li values (− 21.0‰ to + 20.2‰) (Table 1S; Fig. 3). They overlap with the results of 0.20 ppm to 2.40 ppm and − 22.4‰ to + 8.3‰ reported by Tang et al. (2007), but span a broader range of compositions. The ranges of average Li concentrations and δ7Li values are 0.85 ppm to 10.6 ppm and + 3.7‰ to +15.2‰, respectively. The cores of the Opx generally have higher Li concentrations than the rims, but δ7Li variations from core to rim are variable among grains (Fig. 2S). Lithium concentrations in the Hannuoba Cpx span a wide range from 1.18 ppm to 79.8 ppm with 3.72 ppm to 44.8 ppm on average, and the spongy rims of these Cpx, which surround the well-defined Cpx core and have been attributed to decompression melting (Su et al., 2011), have the lowest Li concentrations of 1.18 ppm to 6.02 ppm (Table 1S). The δ7Li values vary from − 17.4‰ to + 18.9‰ with − 5.7‰ to +13.2‰ on average, overlapping with and extending beyond the results of − 0.1‰ to − 10.3‰ reported by Tang et al. (2007). Many of the grains have lower Li contents and δ7Li values in the core compared to the rim and display the same systematic distribution pattern for Li contents and δ7Li from core to rim (Fig. 3S). For individual samples from Hannuoba, Li concentrations vary in the order of Cpx N Ol N Opx. As for Li isotopic compositions, the δ7Li in Ol is overall higher than in Opx and Cpx (Table 1S; Figs. 4, 5). 4.2. Haoti xenoliths Lithium and its isotopic compositions in the Haoti xenoliths also exhibit inter- and intra-mineral heterogeneities (Su et al., 2012b). Lithium isotopic zonation is observed in all minerals. The Ol in these xenoliths have Li concentrations between 1.23 ppm and 13.2 ppm (except two spots up to 30 ppm), and display a large δ7Li variation from − 29.1‰ to + 19.9‰ (Table 1S). The average Li concentration and δ7Li are in the range of 1.53 ppm to 22.1 ppm and −16.4‰ to +14.7‰, respectively. For individual grains, Li concentration and δ7Li are positively correlated, with relatively lower Li in the cores than in the rims (Fig. 1S). The Opx also display apparent Li compositional zonation. They have high and variable Li concentrations (3.00 ppm to 82.8 ppm) and heterogeneous Li isotopic composition, with δ7Li varying from − 16.9‰ to + 18.0‰ (Table 1S; Fig. 3). Within individual grains, Li concentrations decrease from core to rim while δ7Li values increase (Fig. 2S). The Cpx displays remarkably large ranges of Li concentrations (1.39 ppm to 112 ppm) and δ 7Li values (− 45.1‰ to + 19.6‰) (Fig. 3). Lithium concentration and δ7Li value in individual grains are positively correlated (Fig. 3S). The spongy rims of the Cpx show considerably lower Li concentrations but almost identical δ7Li values to those in the core and rim (Table 1S). In Haoti samples, the Cpx has overall higher Li concentrations than Opx and Ol. This pattern of Li distribution is different from the equilibrated Li partitioning sequence established by Seitz and Woodland (2000). Although the δ7Li displays a complicated variation among different minerals in each individual sample, the Ol has lower average δ7Li values (Fig. 3). The average Li concentrations of minerals from both Hannuoba and Haoti xenoliths are higher compared to most of published data. The Hannuoba xenoliths have an average of 2–5 ppm Li in Ol and up to 10 ppm in Opx. The Haoti Ol has Li concentrations of N 5 ppm, except for three samples, while the Haoti pyroxenes exhibit overall more Li enrichment than the Hannuoba samples (Fig. 4).

Fig. 4. Li content vs. Li isotope of individual spot analyses in Ol, Opx and Cpx from the Hannuoba and Haoti xenoliths compared to worldwide data. Data of two Haoti samples in Su et al. (2012b) are also plotted. In figure c, the data outside the defined Haoti field are from spongy rims of Cpx, which resulted from decompression melting with no affinity with metasomatic agents (Su et al., 2011). The results of minerals from Hannuoba peridotites determined on MC-ICP-MS and SIMS (Tang et al., 2007, 2010) are also plotted here for comparison. Lithium data from worldwide localities are the same as in Fig. 3.

5. Discussion Compared to the published Li elemental and isotopic data of mantle minerals worldwide, minerals of the metasomatized xenoliths in this study display apparent Li enrichment and large Li isotopic variations (Fig. 4). Particularly, Li concentrations of Ol, Opx and Cpx in the Haoti xenoliths are as high as 32 ppm, 90 ppm and 110 ppm, respectively (Fig. 4), which are the highest values among all the published data for mantle minerals. The δ7Li values of Ol in these two suites of metasomatized xenoliths extend from +42‰ to −42‰, while the pyroxenes show similar Li isotopic variations to those previously reported (Fig. 4). The lithospheric mantle is thus more heterogeneous in Li abundances and isotopic compositions than previously expected (Fig. 3). Mantle metasomatism in the lithospheric mantle plays a key role in modifying the Li elemental and isotopic compositions in these peridotite phases.

72

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

Fig. 5. a, b, c) Li–Li diagrams showing average Li concentrations in coexisting Ol, Cpx and Opx from Hannuoba and Haoti xenoliths. Equilibrium partitioning (W) line, carbonatite metasomatism (Carb. Metas. (W)) and silicate metasomatism trend (Silic. Metas. (W)) are after from Woodland et al. (2004). d, e, f) δ7Li–δ7Li diagrams showing average Li isotopic compositions in coexisting Ol, Cpx and Opx from the Hannuoba and Haoti xenoliths. Mantle δ7Li values (Ol: +4.7‰; Opx: +1.1‰; Cpx: −1.1‰) are average of worldwide data (see details in main text). The data plotted here are average values from each sample. The gray arrows in these figures are interpretations from the data collected in this study. These diagrams demonstrate that δ7Li is effective in discriminating silicate and carbonatite metasomatism (Carb. Metas.).

5.1. Li elemental and isotopic discrimination of silicate and carbonatite metasomatism Seitz and Woodland (2000) found a general equilibrium Li partitioning relationship in the mantle minerals as follows: Ol N Cpx N Opx. They suggested that disequilibrium Li distribution among peridotite minerals depends on the type of metasomatic agents involved. Further investigations revealed that Li is more readily being incorporated into Cpx in silicate metasomatism and into Ol in carbonatite metasomatism, and indicated that the inter-phase distribution of Li can help to discriminate different metasomatic agents (Woodland et al., 2004; Tang et al., 2007; Zhang et al., 2010). In the correlation diagrams of Li concentration between coexisting Ol, Opx and Cpx, all Hannuoba xenoliths plot within the silicate metasomatism field (Fig. 5a, b, c), consistent with their silicate metasomatized features described above. The Haoti data are scattered and not clustered within the carbonatite metasomatism field (Fig. 5a, b, c). Similar to Li concentrations, we find, for the first time, that δ7Li values are distinct between Ol from Hannuoba and Ol from Haoti (Fig. 4a). The Ol in silicate-metasomatized Hannuoba xenoliths has higher average δ7Li values (N +6‰) relative to normal mantle, whereas the Ol in carbonatite-metasomatized Haoti samples has lower average δ7Li values (mainly b 0‰) (Fig. 5d). The average δ7Li values of the coexisting Cpx and Opx in the Hannuoba and Haoti samples are higher than typical mantle values and display a positive correlation (Fig. 5e). The Opx in the Hannuoba samples has average δ7Li values of N+ 7‰, whereas that in the Haoti samples is b +9‰ (Fig. 5f). Our data thus suggest that the distinct Li isotopic compositions of Ol and Opx are more sensitive parameters in distinguishing between silicate and carbonatite metasomatic agents than Li concentrations.

5.2. Possible processes for Li differences in metasomatized xenoliths In this section, we integrate previous empirical observations and experimental results with our data to discuss possible processes responsible for different geochemical features of Li isotopes relating to silicate and carbonatite metasomatism including: 1) equilibrium isotope fractionation; 2) interaction with host magma; 3) mantle metasomatism; and 4) distinct geochemical behaviors of Li in different metasomatic agents. Here we prefer mantle metasomatism and highlight geochemical behavior of Li. 5.2.1. Equilibrium isotope fractionation, interaction with host magma or mantle metasomatism? The inter-mineral Li distribution and intra-grain Li profiles in the two suites of mantle xenoliths studied here (Figs. 1S, 2S, 3S) indicate that they do not reach isotopic equilibrium, which is further supported by the absence of correlations between their extremely large Li concentrations and Li isotopic variations (Jeffcoate et al., 2007; Wagner and Deloule, 2007). Considering that published data for carbonatites and mantle-derived mafic magmas have similar Li isotopic compositions (δ7Li: 0‰ to +7‰) compared to those of unmetasomatized mantle peridotites (e.g., Ryan and Kyle, 2004; Magna et al., 2006; Halama et al., 2008; Tomascak et al., 2008; Halama et al., 2009), the distinct δ7Li values are thus not likely inherited from metasomatic agents. Several studies have suggested that xenolith–host magma interaction during xenolith transport could also cause Li isotopic variations with several ‰ shifts in δ7Li (Parkinson et al., 2007; Tang et al., 2007; Ionov and Seitz, 2008). In this scenario, the two types of metasomatized xenoliths should exhibit the same variability in δ7Li. This may be true if the host magmas have the same Li concentration and δ7Li, but is not the

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

case for the host magmas of xenoliths in this study. The kamafugites from the Western Qinling have Li concentrations of 9.88 ppm to 26.7 ppm and δ7Li values of + 3.5‰ to + 9.0‰, which overlap the worldwide basalt field (Su et al., 2012b). Furthermore, the observed large and distinctive Li isotopic variations, particularly the opposite δ7Li trends in Ol in silicate and carbonatite metasomatized samples cannot be explained by xenolith–host magma interaction. Since 6Li diffuses faster than 7Li, then metasomatized mantle minerals are expected to have higher Li concentration accompanied by lower δ7Li values as Li in the melt diffuses into the rock (Lundstrom et al., 2005; Rudnick and Ionov, 2007; Teng et al., 2007; Wagner and Deloule, 2007; Wunder et al., 2007). A negative correlation between Li concentration and δ7Li value is thus expected. However, no such correlation has been observed in the samples studied here, which is consistent with previous studies (e.g., Ryan and Kyle, 2004; Elliott et al., 2006; Tang et al., 2007; Aulbach and Rudnick, 2009; Halama et al., 2009; Tang et al., 2011). Moreover, the in situ analytical profiles on minerals illustrate that diverse intra-grain Li isotopic variation trends could occur within individual samples (Figs. 1S, 2S, 3S). This complicated feature of Li isotopes has been widely observed and interpreted as records of multi-stage metasomatic events although diffusion may somewhat yield locally intra-grain variable profiles (e.g., Tang et al., 2007; Wagner and Deloule, 2007; Mallmann et al., 2009; Zhang et al., 2010; Tang et al., 2011). Therefore, the Li enrichment and distinct Li isotopic characteristics in constituent minerals (especially in Ol) of silicate and carbonatite metasomatized xenoliths (Figs. 4a, 5d, f) are most likely linked to different geochemical behaviors of Li isotopes in these two types of metasomatic melts. 5.2.2. Behaviors of Li isotopes in silicate and carbonatite metasomatism Based on the different diffusion rates between 6Li and 7Li, preferential 6Li incorporation relative to 7Li and the resultant low δ7Li in Cpx are followed by rapid diffusion of Li into Cpx, while the relatively 6Lidepleted melts and slower diffusion of Li in Ol would result in higher

73

δ7Li for Ol. Silicate metasomatism would thus result in low-δ7Li Cpx and high-δ7Li Ol in mantle peridotites (Rudnick and Ionov, 2007; Tang et al., 2007; Ionov and Seitz, 2008), which agrees well with the observed 7 Li enrichment order of δ7LiOl N δ7LiOpx ≥ δ7LiCpx in the silicatemetasomatized xenoliths from Hannuoba (Figs. 4, 5d, e, f; Tang et al., 2007) and worldwide localities (Nishio et al., 2004; Seitz et al., 2004; Rudnick and Ionov, 2007; Wagner and Deloule, 2007; Zhang et al., 2010). However, the above observations and theoretical inferences cannot explain the Li isotopic characteristics in the Haoti xenoliths. The contrasting δ7Li sequence (δ7LiCpx N δ7LiOpx N δ7LiOl) in the Haoti minerals suggests that Li is preferentially incorporated into Ol relative to Cpx from carbonatite melts. The change of Li preferential incorporation is probably related to the distinct process of carbonatite metasomatism compared to the silicate counterpart. Numerous studies have indicated that carbonatite metasomatism will increase CaO contents of Ol and Opx, and Na2O content of Cpx, but insignificant change in MgO and FeO contents in these minerals (Yaxley et al., 1991; Laurora et al., 2001; Neumann et al., 2002; Su et al., 2010a). Compared to the Hannuoba xenoliths, the Haoti Ol displays overall high CaO and MgO contents (Fig. 6) and the Haoti Opx shows higher CaO contents and restricted MgO contents (Table 1S; Fig. 7). The breakdown texture of Opx induced by percolating carbonatite melt is commonly observed in the Haoti peridotites (Su et al., 2010b, 2012c). All these observations indicate that the reaction of Ol and Opx with melt was dominated at the initial stage of carbonatite metasomatism and Li could be preferentially incorporated into Ol and Opx, leading to significant Li enrichment and low δ7Li in both minerals relative to the melt. The continuously crystallizing Cpx during carbonatite metasomatism together with older crystals spans a large Li isotopic range as shown in Fig. 4c. The Opx in the Hannuoba xenoliths was formed in multiple silicate metasomatism agents (e.g., Liu et al., 2005; Tang et al., 2007; Zhang et al., 2009) and is thus highly variable in Li concentration and isotopic compositions, even overlaps those of Haoti xenoliths, although they show somewhat differences in Li concentration.

Fig. 6. Major elements versus Li and δ7Li of individual spot analyses in Ol from the Hannuoba and Haoti xenoliths. Mantle values (Li concentration: 1.79 ppm; δ7Li: +4.7‰) of Ol are the average values of worldwide data shown in Fig. 3 and references therein. The vertical bars represent the mantle values of Li concentration and δ7Li, which are the same in Figs. 7 and 8.

74

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

Fig. 7. Major elements versus Li and δ7Li of individual spot analyses in Opx from the Hannuoba and Haoti xenoliths. Mantle values (Li concentration: 1.62 ppm; δ7Li: +1.1‰) of Opx are the average values of worldwide data shown in Fig. 3 and references therein.

5.3. Possible mechanisms for behaviors of Li during mantle metasomatism 5.3.1. Experimental and empirical constraints Many experimental results on silicate magmas and observations in silicate-metasomatized samples revealed that the diffusion rate of Li from melt to Ol is at least several times slower than that to Cpx (Woodland et al., 2004; Coogan et al., 2005; Jeffcoate et al., 2007; Parkinson et al., 2007). Yakob et al. (2012) proposed that Li redistribution could be in response to diffusive addition/subtraction of Li during interaction with infiltrating melt/fluid. Lithium partition coefficients between mineral and silicate melts are in the order of DOl/melt N DCpx/melt N DOpx/melt (Ottolini et al., 2009), which are consistent with the coefficients obtained from natural mineral pairs (Ottolini et al., 2009; Yakob et al., 2012). It is worth to note that this varied DOl–Cpx value could be changed due to a compositional effect in one or the other phase, or both (Woodland et al., 2004). In natural mantle xenoliths, Li in Ol and Cpx displays a broad array of apparent partition coefficients ranging from ~ 0.2 to 10 (Yakob et al., 2012). Thus, most of the empirical data suggest that the interaction between peridotite and melt would increase Li in Ol and pyroxenes (e.g., Seitz et al., 2004; Jeffcoate et al., 2007; Rudnick and Ionov, 2007; Wagner and Deloule, 2007; Zhang et al., 2010).

5.3.2. Elemental exchange and coupled substitution controls The major elemental substitutions in silicate and carbonatite metasomatism are dominated by Fe–Mg and Ca–Mg exchange, respectively. Iron–Mg exchange in Ol is 4 to 8 times faster than the Li diffusion in Ol and 20 to 30 times slower than Li in Cpx (Parkinson et al., 2007). Dohmen et al. (2010) observed a complex diffusion behavior of Li and proposed a coupled fast and slow diffusion mechanism. The fast Li diffusion is unlikely to be dominant in most natural systems with the exception of Li-rich and oxidizing environments; under slow diffusion, the Li diffusion rate is about an order of magnitude faster than diffusion of Fe, Mg and most other divalent cations in Ol, and is much slower than the

rate in Cpx under the same conditions (Dohmen et al., 2010). The diffusion rate and cation exchange are thus relevant to the element content in the minerals. The faster diffusion rate of Li in Cpx relative to Li and Fe–Mg exchange in Ol would enhance the Li enrichment, with up to tens of ppm in Cpx, if given sufficient reaction time. The overall low MgO content in minerals from the Hannuoba xenoliths (Figs. 6, 7, 8) indicates that silicate metasomatism is dominated by Fe–Mg exchange. By contrast, the relatively higher MgO content and its limited variation in Haoti minerals (Figs. 6, 7, 8) suggest weak Fe–Mg exchange during carbonatite metasomatism, whereas the high CaO contents in Opx reflect significant Ca–Mg exchange. The positive correlation between CaO and Li in the Haoti Opx (Fig. 7) implies that Ca–Mg exchange may facilitate Li incorporation into minerals from carbonatite melt. As for a substitution mechanism for Li from melt into mantle mineral, an experimental study by Wunder et al. (2006) indicated that Li could replace Mg in synthetic Cpx. The coupled substitution of Li and P and Na to maintain charge balance has been emphasized by some authors (Seitz and Woodland, 2000; Wagner and Deloule, 2007; Mallmann et al., 2009). Specifically, a strong correlation of Li and Na with P in Ol has been observed and thus coupled substitutions of Li with P (Li+ + P5+ = Mg2+ + Si4+) are proposed in order to maintain charge balance in Ol (Seitz and Woodland, 2000; Mallmann et al., 2009). Theoretically, since both Li and Na have a + 1 valence, they could be balanced by higher valence elements such as Mg. The overall positive correlation between Na2O and Li contents in Cpx, especially from carbonatite metasomatized samples (Fig. 8), implies that a coupled substitution of Li with Na could significantly promote the incorporation of Li into pyroxenes. Many studies have also observed that Na-enriched melts generally have higher Li contents. Zack et al. (2003) discussed an enrichment of Li due to the formation of chlorite (where Li substitutes for Mg) associated with an increase in Na (due to seafloor spilitization, i.e., the formation of albite). Wagner and Deloule (2007) briefly mentioned diffusive Li–Na exchange between melt and Cpx. Su et al. (2012b) documented that the presence of Na+1 might enhance the incorporation of Li from melt into mineral phases through the

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

75

Fig. 8. Major elements versus Li and δ7Li of individual spot analyses in Cpx from the Hannuoba and Haoti xenoliths. Mantle values (Li concentration: 3.22 ppm; δ7Li: −1.1‰) of Cpx are the average values of worldwide data shown in Fig. 3 and references therein.

coupled substitution of Li+1 and Na+1 for Mg+2. Further supports come from the occurrence of Li-enriched Cpx from nephelinites, alkaline and agpaitic rocks, owing to the similar chemical nature between Li and Na (Halama et al., 2007; Marks et al., 2007; Su et al., 2012b). Therefore, if the reactant/metasomatic melts are enriched in Li and Na, the reaction could explain why the Cpx has high concentrations of both elements. The above inferences could help to partly explain the distinct Li elemental and isotopic distribution in silicate and carbonatite metasomatized samples. 6. Conclusions (1) The newly obtained Li abundances and isotopic data for carbonatite metasomatized Haoti xenoliths from Western Qinling, compared to their counterparts for silicate metasomatized Hannuoba samples from North China Craton, provide strong support for the presence of a highly heterogeneous lithospheric mantle. The striking Li enrichment and large Li isotopic variations can be attributed to mantle metasomatism. (2) The Hannuoba and Haoti xenoliths display apparently distinct Li contents and isotopic compositions with Ol and pyroxenes in the

Haoti xenoliths showing significantly higher Li concentrations than those in the Hannuoba samples. The Haoti xenoliths are characterized by Li isotopic composition following the order of δ7LiCpx N δ7LiOpx N δ7LiOl, whereas the Hannuoba xenoliths have a sequence of δ7LiOl N δ7LiOpx N δ7LiCpx. Lithium isotope signatures can thus help to discriminate silicate and carbonatite metasomatic agents. (3) The distinct geochemical characteristics of Li isotopes in silicate and carbonatite metasomatized samples are probably closely related to the preferential incorporation of Li into minerals from distinct melts. In carbonatite metasomatism, the Opx-consumption and Cpx-formation processes indicate that Opx and Ol are the first mineral phases that interact with the carbonatite melt, generating the low-δ7Li features at the initial stage of carbonatite metasomatism. Acknowledgments We are grateful to Raphaël Pik, Qian Mao and Yu-Guang Ma for their assistance in measurement of Li isotopes and major elements. We thank Fang-Zhen Teng and Penniston-Dorland Sarah for improving the

76

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77

original version of the paper. Paul B. Tomascak, Klaus Mezger and another anonymous reviewer are warmly acknowledged for their careful and constructive reviews, which considerably improved this paper. This work was financially supported by the National Natural Science Foundation of China (Grants 41173011, 91214203, 41173009 and 91014007) and Hong Kong Scholars Program (No. XJ2012048). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.chemgeo.2014.05.016. References Aulbach, S., Rudnick, R.L., 2009. Origins of non-equilibrium lithium isotope fractionation in xenolithic peridotite minerals: examples from Tanzania. Chem. Geol. 258, 17–27. Bell, D.R., Hervig, R.L., Buseck, P.R., Aulbach, S., 2009. Lithium isotope analysis of olivine by SIMS: calibration of a matrix effect and application to magmatic phenocrysts. Chem. Geol. 258, 5–16. Chen, S.H., O'Reilly, S.Y., Zhou, X.H., Griffin, W.L., Zhang, G.H., Sun, M., Feng, J.L., Zhang, M., 2001. Thermal and petrological structure of the lithosphere beneath Hannuoba, SinoKorean Craton, China: evidence from xenoliths. Lithos 56, 267–301. Choi, S.H., Mukasa, S.B., Zhou, X.H., Xian, X.H., Andronikov, A.V., 2008. Mantle dynamics beneath East Asia constrained by Sr, Nd, Pb and Hf isotopic systematics of ultramafic xenoliths and their host basalts from Hannuoba, North China. Chem. Geol. 248, 40–61. Coogan, L.A., Kasemann, S.A., Chakraborty, S., 2005. Rates of hydrothermal cooling of new oceanic upper crust derived from lithium-geospeedometry. Earth Planet. Sci. Lett. 240, 415–424. Dautria, J.M., Dupuy, C., Takherist, D., Dostal, J., 1992. Carbonate metasomatism in the lithospheric mantle: the peridotitic xenoliths from a melilititic district of the Sahara Basin. Contrib. Mineral. Petrol. 111, 37–52. Decitre, S.E., Deloule, E., Reisberg, L., James, R., Agrinier, P., Mevel, C., 2002. Behavior of Li and its isotopes during serpentinization of oceanic peridotites. Geochem. Geophys. Geosyst. 3. http://dx.doi.org/10.1029/2001GC000178. Deloule, E., Chaussidon, M., Alle, P., 1992. Instrumental limitations for isotope measurements with a Cameca ims-3f ion microprobe: example of H, B, S and Sr. Chem. Geol. 101, 187–192. Dobson, D.P., Jones, A.P., Rabe, R., Sekine, T., Kurita, K., Taniguchi, T., Kondo, T., Kato, T., Shimomura, O., Urakawa, S., 1996. In-situ measurement of viscosity and density of carbonate melts at high pressure. Earth Planet. Sci. Lett. 143, 207–215. Dohmen, R., Kasemann, S.A., Coogan, L., Chakraborty, S., 2010. Diffusion of Li in olivine. Part I: experimental observations and a multi species diffusion model. Geochim. Cosmochim. Acta 74, 274–292. Du, X.X., Fan, Q.C., 2011. The genesis of Si-rich melt in mantle xenoliths from Hannuoba and their significance. Acta Petrol. Sin. 27, 1267–1274 (in Chinese with English abstract). Elliott, T., Thomas, A., Jeffcoate, A., Niu, Y.L., 2006. Lithium isotope evidence for subductionenriched mantle in the source of midocean-ridge basalts. Nature 443, 565–568. Flesch, G.D., Anderson, A.R.J., Svec, H.J., 1973. A secondary isotopic standard for 6Li/7Li determinations. Int. J. Mass Spectrom. Ion Phys. 12, 265–272. Frezzotti, M.L., Andersen, T., Neumann, E.R., Simonsen, S.L., 2002. Carbonatite melt–CO2 fluid inclusions in mantle xenoliths from Tenerife, Canary Islands: a story of trapping, immiscibility and fluid–rock interaction in the upper mantle. Lithos 64, 77–96. Gaillard, F., Malki, M., Iacono-Marziano, G., Pichavant, M., Scaillet, B., 2008. Carbonatite melts and electrical conductivity in the asthenosphere. Science 322, 1363–1365. Gasparik, T., Litvin, Y.A., 2002. Experimental investigation of the effect of metasomatism by carbonatic melt on the composition and structure of the deep mantle. Lithos 60, 129–143. Genge, M.J., Price, G.D., Jones, A.P., 1995. Molecular dynamics simulations of CaCO3 melts to mantle pressures and temperatures: implications for carbonatite magmas. Earth Planet. Sci. Lett. 131, 225–238. Gorring, M.L., Kay, S.M., 2000. Carbonatite metasomatized peridotite xenoliths from southern Patagonia: implication for lithospheric processes and Neogene plateau magmatism. Contrib. Mineral. Petrol. 140, 55–72. Green, T.H., Adam, J., Sie, S.H., 1992. Trace element partitioning between silicate minerals and carbonatite at 25 kbar and application to mantle metasomatism. Mineral. Petrol. 46, 179–184. Halama, R., McDonough, W.F., Rudnick, R.L., Keller, J., Klaudius, J., 2007. The Li isotopic composition of Oldoinyo Lengai: nature of the mantle sources and lack of isotopic fractionation during carbonatite petrogenesis. Earth Planet. Sci. Lett. 254, 77–89. Halama, R., McDonough, W.F., Rudnick, R.L., Bell, K., 2008. Tracking the lithium isotopic evolution of the mantle using carbonatites. Earth Planet. Sci. Lett. 265, 726–742. Halama, R., Savov, I.P., Rudnick, R.L., McDonough, W.F., 2009. Insights into Li and Li isotope cycling and sub-arc metasomatism from veined mantle xenoliths, Kamchatka. Contrib. Mineral. Petrol. 158, 197–222. Ionov, D.A., Seitz, H.M., 2008. Lithium abundances and isotopic compositions in mantle xenoliths from subduction and intra-plate settings: mantle sources vs. eruption histories. Earth Planet. Sci. Lett. 266, 316–331. Ionov, D.A., O'Reilly, S.Y., Genshaft, Y.S., Kopylova, M.G., 1996. Carbonate-bearing mantle peridotite xenoliths from Spitsbergen: phase relationships, mineral compositions and trace-element residence. Contrib. Mineral. Petrol. 125, 375–392.

Jeffcoate, A.B., Elliott, T., Kasemann, S.A., Ionov, D., Cooper, K., Brooker, R., 2007. Li isotope fractionation in peridotites and mafic melts. Geochim. Cosmochim. Acta 71, 202–218. Klemme, S., Van der Laan, S.R., Foley, S.F., Günther, D., 1995. Experimentally determined trace and minor element partitioning between clinopyroxene and carbonatite melt under upper mantle conditions. Earth and Planetary Science Letters 133, 439–448. Laurora, A., Mazzucchelli, M., Rivalenti, G., Vannucci, R., Zanetti, A., Barbieri, M.A., Cingolani, C.A., 2001. Metasomatism and melting in carbonated peridotite xenoliths from the mantle wedge: the Gobernator Gregores case (Southern Patagonia). J. Petrol. 42, 69–87. Liu, R., Chen, W., Sun, J., Li, D., 1990. K–Ar age and tectonic environment of Cenozoic volcanic rock in China. In: Liu, R. (Ed.), Age and Geochemistry of Cenozoic Volcanic Rock in China. Seismology Published House, Beijing, pp. 1–43 (in Chinese). Liu, Y.S., Gao, S., Liu, X.M., Chen, X.M., Zhang, W.L., Wang, X.C., 2003. Thermodynamic evolution of lithosphere of the North China craton: records from lower crust and upper mantle xenoliths from Hannuoba. Chin. Sci. Bull. 48, 2371–2377. Liu, Y.S., Yuan, H.L., Gao, S., Hu, Z.C., Wang, X.C., Liu, X.M., Lin, W.L., 2004. Zircon U–Pb ages of olivine pyroxenite xenolith from Hannuoba: links between the 97–158 Ma basaltic underplating and granulite-facies metamorphism. Chin. Sci. Bull. 49, 1055–1062. Liu, Y.S., Gao, S., Lee, C.T.A., Hu, S.Y., Liu, X.M., Yuan, H.L., 2005. Melt–peridotite interactions: links between garnet pyroxenite and high-Mg# signature of continental crust. Earth Planet. Sci. Lett. 234, 39–57. Lundstrom, C.C., Chaussidon, M., Hsui, A.T., Kelemen, P., Zimmerman, M., 2005. Observations of Li isotopic variations in the Trinity ophiolite: evidence for isotopic fractionation by diffusion during mantle melting. Geochim. Cosmochim. Acta 69, 735–751. Magna, T., Wiechert, U., Halliday, A.N., 2006. New constraints on the lithium isotope compositions of the moon and terrestrial planets. Earth Planet. Sci. Lett. 243, 336–353. Mallmann, G., O'Neill, H.St.C., Klemme, S., 2009. Heterogeneous distribution of phosphorus in olivine from otherwise well-equilibrated spinel peridotite xenoliths and its implications for the mantle geochemistry of lithium. Contrib. Mineral. Petrol. 158, 485–504. Marks, M.A.W., Rudnick, R.L., McCammon, C., Vennemann, T., Markl, G., 2007. Arrested kinetic Li isotope fractionation at the margin of the Ilimaussaq complex, South Greenland: evidence for open-system processes during final cooling of peralkaline igneous rocks. Chem. Geol. 246, 207–230. Neumann, E.R., Wulff-Pedersen, E., Pearson, N.J., Spencer, E.A., 2002. Mantle xenoliths from Tenerife (Canary Islands): evidence for reactions between mantle peridotites and silicic carbonatite melts inducing Ca metasomatism. J. Petrol. 43, 825–857. Nishio, Y., Shunichi, N., Yamamoto, J., Sumino, H., Matsumoto, T., Prikhodko, V.S., Arai, S., 2004. Lithium isotopic systematics of the mantle-derived ultramafic xenoliths: implications for EM1 origin. Earth Planet. Sci. Lett. 217, 245–261. Ottolini, L., Laporte, D., Raffone, N., Devidal, J.L., Le Fevre, B., 2009. New experimental determination of Li and B partition coefficients during upper mantle partial melting. Contrib. Mineral. Petrol. 157, 313–325. Parkinson, I.J., Hammond, S.J., James, R.H., Rogers, N.W., 2007. High-temperature lithium isotope fractionation: Insights from lithium isotope diffusion in magmatic systems. Earth and Planetary Science Letters 257, 609–621. Roden, M.F., Murthy, V.R., 1985. Mantle metasomatism. Annu. Rev. Earth Planet. Sci. 13, 169–296. Rudnick, R.L., Ionov, D.A., 2007. Lithium elemental and isotopic disequilibrium in minerals from peridotite xenoliths from far-east Russia: product of recent melt/fluid–rock reaction. Earth Planet. Sci. Lett. 256, 278–293. Ryan, J.G., Kyle, P.R., 2004. Lithium abundance and lithium isotope variations in mantle sources: insights from intraplate volcanic rocks from Ross Island and Marie Byrd Land (Antarctica) and other oceanic islands. Chem. Geol. 212, 125–142. Seitz, H.M., Woodland, A.B., 2000. The distribution of lithium in peridotitic and pyroxenitic mantle lithologies — an indicator of magmatic and metasomatic processes. Chem. Geol. 166, 47–64. Seitz, H.M., Brey, G.P., Lahaye, Y., Durali, S., Weyer, S., 2004. Lithium isotopic signatures of peridotite xenoliths and isotopic fractionation at high temperature between olivine and pyroxenes. Chem. Geol. 212, 163–177. Song, Y., Frey, F.A., 1989. Geochemistry of peridotite xenoliths in basalt from Hannuoba, eastern China: implications for subcontinental mantle heterogeneity. Geochim. Cosmochim. Acta 53, 97–113. Su, B.X., Zhang, H.F., Ying, J.F., Xiao, Y., Zhao, X.M., 2009. Nature and processes of the lithospheric mantle beneath the western Qinling: evidence from deformed peridotitic xenoliths in Cenozoic kamafugite from Haoti, Gansu Province, China. J. Asian Earth Sci. 34, 258–274. Su, B.X., Zhang, H.F., Sakyi, P.A., Ying, J.F., Tang, Y.J., Yang, Y.H., Qin, K.Z., Xiao, Y., Zhao, X.M., 2010a. Compositionally stratified lithosphere and carbonatite metasomatism recorded in mantle xenoliths from the Western Qinling (Central China). Lithos 116, 111–128. Su, B.X., Zhang, H.F., Sakyi, P.A., Qin, K.Z., Liu, P.P., Ying, J.F., Tang, Y.J., Malaviarachchi, S.P.K., Xiao, Y., Zhao, X.M., Mao, Q., Ma, Y.G., 2010b. Formation of melt pockets in mantle peridotite xenoliths from the Western Qinling (Central China): partial melting and metasomatism. J. Earth Sci. 21, 641–668. Su, B.X., Zhang, H.F., Sakyi, P.A., Yang, Y.H., Ying, J.F., Tang, Y.J., Qin, K.Z., Xiao, Y., Zhao, X.M., Mao, Q., Ma, Y.G., 2011. The origin of spongy texture of mantle xenolith minerals from the Western Qinling, Central China. Contrib. Mineral. Petrol. 161, 465–482. Su, B.X., Zhang, H.F., Ying, J.F., Tang, Y.J., Hu, Y., Santosh, M., 2012a. Metasomatized lithospheric mantle beneath Western Qinling, central China: insight into activity of carbonatite melts. J. Geol. 120, 671–681. Su, B.X., Zhang, H.F., Deloule, E., Sakyi, P.A., Xiao, Y., Tang, Y.J., Hu, Y., Ying, J.F., Liu, P.P., 2012b. Extremely high Li and low δ7Li signatures in the lithospheric mantle. Chem. Geol. 292–293, 149–157.

B.-X. Su et al. / Chemical Geology 381 (2014) 67–77 Su, B.X., Zhang, H.F., Yang, Y.H., Sakyi, P.A., Ying, J.F., Tang, Y.J., 2012c. Breakdown of orthopyroxene contributing to melt pockets in mantle peridotite xenoliths from the Western Qinling, central China: constraints from in situ LA-ICP-MS mineral analyses. Mineral. Petrol. 104, 225–247. Sweeney, R.J., Prozesky, V., Przybylowicz, W., 1995. Selected trace and minor element partitioning between peridotite minerals and carbonaitire melts at 18–46 kbar pressure. Geochim. Cosmochim. Acta 59, 3671–3683. Tang, Y.J., Zhang, H.F., Nakamura, E., Moriguti, T., Kobayashi, K., Ying, J.F., 2007. Lithium isotopic systematics of peridotite xenoliths from Hannuoba, North China Craton: implications for melt/rock interaction in the considerably thinned lithospheric mantle. Geochim. Cosmochim. Acta 71, 4327–4341. Tang, Y.J., Zhang, H.F., Ying, J.F., 2010. A brief review of isotopically light Li—a feature of the enriched mantle? Int. Geol. Rev. 52, 964–976. Tang, Y.J., Zhang, H.F., Nakamura, E., Ying, J.F., 2011. Multistage melt/fluid-peridotite interactions in the refertilized lithospheric mantle beneath the North China Craton: constraints from the Li–Sr–Nd isotopic disequilibrium between minerals of peridotite xenoliths. Contrib. Mineral. Petrol. 161, 845–861. Tang, Y.J., Zhang, H.F., Deloule, E., Su, B.X., Ying, J.F., Xiao, Y., Hu, Y., 2012. Slab-derived lithium isotopic signatures in mantle xenoliths from northeastern North China Craton. Lithos 149, 79–90. Teng, F.Z., McDonough, W.F., Rudnick, R.L., Wing, B.A., 2007. Limited lithium isotopic fractionation during progressive metamorphic dehydration in metapelites: a case study from the Onawa contact aureole, Maine. Chem. Geol. 239, 1–12. Tomascak, P.B., Langmuir, C.H., le Roux, P., Shirey, S.B., 2008. Lithium isotopes in global mid-ocean ridge basalts. Geochim. Cosmochim. Acta 72, 1626–1637. Veksler, V., Petibon, C., Jenner, J.A., Dorfman, A.M., Dingwell, D.B., 1998. Trace element partitioning in immiscible silicate–carbonate systems: an initial experimental study using a centrifuge auto clave. J. Petrol. 39, 2095–2104. Wagner, C., Deloule, E., 2007. Behaviour of Li and its isotopes during metasomatism of French Massif Central lherzolites. Geochim. Cosmochim. Acta 71, 4279–4296. Woodland, A.B., Seitz, H.M., Yaxley, G.M., 2004. Varying behaviour of Li in metasomatised spinel peridotite xenoliths from western Victoria, Australia. Lithos 75, 55–66. Wunder, B., Meixner, A., Romer, R.L., Heinrich, W., 2006. Temperature-dependent isotopic fractionation of lithium between clinopyroxene and high pressure hydrous fluids. Contributions to Mineralogy and Petrology 151, 112–120.

77

Wunder, B., Meixner, A., Romer, R.L., Feenstra, A., Schettler, G., Heinrich, W., 2007. Lithium isotope fractionation between Li-bearing staurolite, Li-mica and aqueous fluids: an experimental study. Chem. Geol. 238, 277–290. Yakob, J.L., Feineman, M.D., Deane Jr., J.A., Eggler, D.H., Penniston-Dorland, S.C., 2012. Lithium partitioning between olivine and diopside at upper mantle conditions: an experimental study. Earth Planet. Sci. Lett. 329–330, 11–21. Yaxley, G.M., Crawford, A.J., Green, D.H., 1991. Evidence for carbonatite metasomatism in spinel peridotite xenoliths from western Victoria, Australia. Earth Planet. Sci. Lett. 107, 305–317. Ying, J.F., Zhou, X.H., Zhang, H.F., 2004. Geochemical and isotopic investigation of the Laiwu–Zibo carbonatites from western Shandong Province, China, and implications for their petrogenesis and enriched mantle source. Lithos 75, 413–426. Yu, X.H., Zhao, Z.D., Mo, X.X., Zhou, S., Zhu, D.Q., Wang, Y.L., 2005. 40Ar/39Ar dating for Cenozoic kamafugites from western Qinling in Gansu Province. Chin. Sci. Bull. 50, 2638–2643. Zack, T., Tomascak, P.B., Rudnick, R.L., Dalpe, C., McDonough, W.F., 2003. Extremely light Li in orogenic eclogites: the role of isotope fractionation during dehydration in subducted oceanic crust. Earth Planet. Sci. Lett. 208, 279–290. Zhang, H.F., Goldstein, S.L., Zhou, X.H., Sun, M., Cai, Y., 2009. Comprehensive refertilization of lithospheric mantle beneath the North China Craton: further Os–Sr–Nd isotopic constraints. J. Geol. Soc. Lond. 166, 249–259. Zhang, H.F., Deloule, E., Tang, Y.J., Ying, J.F., 2010. Melt/rock interaction in remains of refertilized Archean lithospheric mantle in Jiaodong Peninsula, North China Craton: Li isotopic evidence. Contrib. Mineral. Petrol. 160, 261–277. Zhang, H.F., Zhu, R.X., Santosh, M., Ying, J.F., Su, B.X., Hu, Y., 2013. Episodic widespread magma underplating beneath the North China Craton in the Phanerozoic: implications for craton destruction. Gondwana Res. 23, 95–107. Zheng, J.P., Griffin, W.L., Qi, L., O'Reilly, S.Y., Sun, M., Zheng, S., Pearson, N., Gao, J.F., Yu, C.M., Su, Y.P., Tang, H.Y., Liu, Q.S., Wu, X.L., 2009. Age and composition of granulite and pyroxenite xenoliths in Hannuoba basalts reflect Paleogene underplating beneath the North China Craton. Chem. Geol. 264, 266–280.