Lithos 286–287 (2017) 363–368
Contents lists available at ScienceDirect
Lithos
j ourna l homepage: www.e lsev ie r .com/ locate / l i thosLithium isotopic composition of Alaskan-type intrusion and
its implicationBen-Xun Su a,b,⁎, Chen Chen a,b, Yang Bai a,b, Kwan-Nang Pang c, Ke-Zhang Qin a,b, Patrick Asamoah Sakyi d,⁎⁎
a Key Laboratory of Mineral Resources, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China
b University of Chinese Academy of Sciences, Beijing 100049, China
c Institute of Earth Sciences, Academia Sinica, Taipei 11529, Taiwan
d Department of Earth Science, School of Physical and Mathematical Sciences, University of Ghana, P.O. Box LG 58, Legon-Accra, Ghana⁎ Correspondence to: B.-X. Su, Key Laboratory ofMinera
and Geophysics, Chinese Academy of Sciences, Beijing 100
⁎⁎ Corresponding author.
E-mail addresses: subenxun@mail.igcas.ac.cn (B.-X. Su
(P.A. Sakyi).
http://dx.doi.org/10.1016/j.lithos.2017.06.024
0024-4937/© 2017 Elsevier B.V. All rights reserved.a b s t r a c ta r t i c l e i n f oArticle history:
Received 27 February 2017
Accepted 22 June 2017
Available online 28 June 2017To determinewhether fractionation of Li isotopes could occur at high temperatures, we examinedmajor element
and Li isotopic compositions of olivine from the Xiadong intrusion, an Alaskan-type complex in the Central Asian
Orogenic Belt. Olivine in thirteen dunites, displaying characteristic cumulus textures, yielded large variations in Li
concentration (0.10 to 11.18 ppm) and isotopic composition (δ7Li = −7.18 to +34.41‰). These variations are
too large to be attributed entirely to diffusive processes. The correlations between Li elemental or isotopic com-
position and differentiation indices such as Fo andMnO contents of olivine, and NiO content of chromite, suggest
probable Li isotope fractionation during early stage of differentiation. We speculate that while Li behaves mildly
incompatible during differentiation, 7Li is preferentially incorporated into olivine relative to 6Li during early stage
crystallization from arcmagmas. Relatively high Li concentrations and low δ7Li in arcmagmasmight be the result
of substantial olivine fractionation prior to eruption.
© 2017 Elsevier B.V. All rights reserved.Keywords:
Li isotopes
Alaskan-type intrusion
Arc magmatism
Olivine
Magma differentiation1. Introduction
A general principle in stable isotope geochemistry is that fraction-
ation occurring at mantle or magmatic temperatures tends to be small
as the fractionation decreases with the square of the temperature
(Chacko et al., 2001; White, 2015). This is also the case for the stable
Li isotopes, 6Li and 7Li, which, in several intra-plate oceanic volcanic
suites, show no measurable variations (Chan and Frey, 2003;
Tomascak et al., 1999). The implications for this are that magmatic
rocks whose Li isotopes show significant variations from the mantle
might have either been affected by low-temperature processes or prob-
ably contain a component that once occurred on the surface of the Earth.
The latter inference represents an important assumption in using Li iso-
topes as a tracer of crustal recycling (see reviews in Chapter 6 by
Tomascak et al., 2016). If Li isotope fractionation during differentiation
is demonstrated, its effect andmagnitude should be quantified to refine
the previous interpretation based on the absence of Li fractionation dur-
ing high-T processes.l Resources, Institute of Geology
029, China.
), pasakyi@ug.edu.ghTomascak et al. (1999) and Chan and Frey (2003) chose intra-plate
oceanic volcanic suites to study the effect of basalt differentiation to
avoid contributions of slab-derived Li and Li from the continental
crust. The approach used was to examine the Li isotopic variations in
bulk-rock samples obtained in solution by multi-collector inductively-
coupled plasma mass spectrometry (MC-ICP-MS), with the assumption
that all samples evolved along a common liquid line of descent. An ad-
ditional way was to compare Li isotopic data of phenocrysts and bulk
matrix separated from a given sample (Chan and Frey, 2003). Both stud-
ies concluded that Li isotopes do not fractionate during basalt differen-
tiation. However, such results obtained from intra-plate volcanic suites
might not be equally applied to basaltic magmas in different tectonic
settings (Ackerman et al., 2015; Marks et al., 2007; Schuessler et al.,
2009; Teng et al., 2006). Several recent investigations have highlighted
the possibility of high-temperature fractionation of Li isotopes. For ex-
ample, the study of Weyer and Seitz (2012) revealed substantial vari-
ability in Li concentrations and isotopic compositions between olivine
and basaltic matrix, implying that Li isotopes might fractionate at mag-
matic temperatures. More recently, Li isotope fractionation has been
observed in mare basalts and was attributed to large-degree, high-
temperature igneous differentiation of their source regions (Day et al.,
2016).
The capability to analyze olivine and pyroxene grains for Li isotopes
by secondary ionmass spectrometry (SIMS)with high spatial resolution
allows amore detailed examination of isotope fractionation down to the
364 B.-X. Su et al. / Lithos 286–287 (2017) 363–368grain scale. This, therefore, serves as themotivation for a study of Li iso-
topes in olivine from Alaskan-type intrusive rocks that formed by hy-
drous basaltic magmas in arc roots (Irvine, 1974; Taylor, 1967). If
magma differentiation does not fractionate Li isotopes, then we should
expect similar Li isotopic compositions between arc magmas and their
early cumulates that are complementary to differentiation. Here, we re-
port the Li isotopic compositions of olivine grains from 12 chromite-
bearing dunite samples with characteristic cumulus texture from the
Xiadong Alaskan-type intrusion in the Central Asian Orogenic Belt
(CAOB). Like other Alaskan-type intrusions worldwide, the Xiadong in-
trusionwas solidified from a hydrous arc-related parentalmagmawith-
out significant crustal contamination (Su et al., 2014). It is highly
fractionated in lithology with simple mineral assemblage in individual
rock types (Su et al., 2012). These features make the Xiadong olivine
an ideal tool to investigate Li isotope fractionation duringmagmadiffer-
entiation and evolution of arc magma at sub-volcanic depth.
2. Samples and methods
The Xiadong intrusion is an Alaskan-type intrusion in the southern
margin of the CAOB and consists of dunite, clinopyroxenite,
hornblendite and gabbro (Su et al., 2012). The dunites are made up of
olivine (80–95 vol.%) and chromite (5–20 vol.%) and show cumulus tex-
ture. The samples selected for Li isotope analysis are mostly fresh as
shown in Fig. 1a, but one of them (09XDTC1-35) is highly serpentinized
and only relics of olivine and chromite are present (Fig. 1b). The olivine
in these dunites displays large chemical variations in Fo (92.3–96.6) and
NiO (0.05–0.76 wt.%) (Bai et al., 2017; Su et al., 2012; Sun et al., 2009).
Chromite is compositionally Fe3+-bearing chromite or Cr-magnetite
and follows a differentiation (Fe enrichment) trend from an intermedi-
ate Cr-Al-rich spinel to Cr-magnetite. The NiO and MnO contents of theFig. 1. Photomicrographs of the dunites from the Xiadong mafic-ultramafic intrusion. (a) Fr
chromite (Chr) and chromite inclusion in olivine. (b) Preserved olivine and chromite relic
09XDTC1-32 showing spot analysis locations (red ellipse).chromite range from 0.37 to1.02 wt.% and 0.05 to 1.18 wt.%, respective-
ly. These dunites, having high MgO contents (39.7–44.5 wt.%), were
considered as ultramafic cumulates of primitive arc magmas (Su et al.,
2014).
Thin sections of the samples were gold-coated for in-situ Li isotope
analyses using a Cameca IMS-1280HR in the Institute of Geology and
Geophysics, Chinese Academyof Sciences, Beijing, China, following sim-
ilar methods as in Decitre et al. (2002). Because the studied olivine is
fine-grained, we conducted only one analysis on each grain. The work-
ing conditions are described in Su et al. (2015, 2016). Olivine samples
06JY-34Ol and 06JY-31Ol (Su et al., 2015) were used as standards and
yielded homogeneous Li isotopic compositions. Matrix effect, of which
δ7Li increased by 1.0‰ for eachmole percent decrease in the Fo of oliv-
ine (Su et al., 2015), was considered for calibration. The Li isotopic com-
positions of olivine separates from two samples (09XDTC1-24 and
09XDTC1-25) measured by solution MC-ICP-MS at the Centre de
Recherches Petrographiques et Geochimiques, Nancy, France (Vigier
et al., 2008), are also reported here. Notably, 09XDTC1-24Ol has been
developed as standard for in situ Li isotope analysis owing to its compo-
sitional homogeneity (Su et al., 2015).
3. Results
Lithium isotopic compositions of olivine and selected geochemical
parameters of the dunites are listed in Table 1. Generally, the Li concen-
trations and δ7Li values of the olivine show large variations from 0.10 to
11.18 ppm and−7.18 to+34.4‰, respectively, and they are negatively
correlated (Fig. 2). The olivine grains are relatively homogeneous in Li
concentration, but display variable Li isotopes in individual samples
(Table 1). The δ7Li value of sample 09XDTC1-25 measured by MC-ICP-
MS falls within the range for olivine from the same sample analyzedesh sample 09XD-1 consisting of olivine (Ol) with well-developed fractures, interstitial
s in highly altered sample 09XDTC1-35. (c) Back-scattered electron image of sample
B.-X. Su et al. / Lithos 286–287 (2017) 363–368 365
Table 1 Table 1 (continued)
Li elemental and isotopic compositions of olivine and selected geochemical parameters in
7 a a
the dunites from the Xiadong Alaskan-type intrusion. Sample@grain δ Li 1se Li 1se Fo in MnO in NiO in
Olb Chrb Chrb
Sample@grain δ7Li 1sea Li 1sea Fo in MnO in NiO in
b b b ‰ ppm wt.% wt.%Ol Chr Chr
Average 8.92 0.30
‰ ppm wt.% wt.%
09XDTC1-35@1 22.06 2.85 0.16 0.00 95.0 0.05 0.37
MC-ICP-MS 09XDTC1-35@2 20.55 1.84 0.17 0.00
09XDTC1-24 8.91 0.11 1.49 0.03 94.2 0.39 1.01 09XDTC1-35@3 34.41 2.39 0.10 0.00
09XDTC1-25 13.10 0.11 1.61 0.06 96.4 0.22 0.96 09XDTC1-35@4 23.80 2.17 0.12 0.00
SIMS 09XDTC1-35@5 23.38 2.04 0.14 0.00
09XDTC1-16@1 3.91 1.32 1.96 0.01 95.1 0.25 0.68 09XDTC1-35@6 23.37 2.06 0.13 0.00
09XDTC1-16@2 9.18 1.46 2.71 0.03 09XDTC1-35@7 25.90 2.09 0.13 0.00
09XDTC1-16@3 13.27 1.02 2.84 0.03 09XDTC1-35@8 27.58 2.30 0.20 0.00
09XDTC1-16@4 5.50 0.49 1.53 0.01 Average 25.13 0.14
09XDTC1-16@5 7.98 0.53 1.48 0.01 09XDTC1-36@1 13.59 1.04 0.61 0.01 95.9 0.52 0.73
09XDTC1-16@6 −0.24 0.60 2.29 0.02 09XDTC1-36@2 5.06 0.63 1.49 0.01
09XDTC1-16@7 1.97 0.93 1.47 0.02 09XDTC1-36@3 16.03 1.38 0.44 0.00
Average 5.94 2.04 09XDTC1-36@4 7.09 0.67 1.25 0.01
09XDTC1-15@1 4.65 0.53 2.19 0.03 94.5 0.41 0.56 09XDTC1-36@5 8.99 0.82 1.00 0.00
09XDTC1-15@2 2.89 0.44 3.06 0.03 09XDTC1-36@6 4.09 0.85 1.27 0.01
09XDTC1-15@3 3.03 0.43 2.80 0.02 09XDTC1-36@7 8.83 0.86 0.87 0.01
09XDTC1-15@4 5.64 0.43 2.72 0.02 09XDTC1-36@8 8.60 0.89 0.80 0.01
09XDTC1-15@5 4.90 0.59 1.98 0.02 Average 9.04 0.97
09XDTC1-15@6 9.20 0.56 1.46 0.01 09XDTC1-29@1 19.47 2.37 0.12 0.00 96.0 0.26 0.86
Average 5.05 2.37 09XDTC1-29@2 18.05 2.54 0.33 0.00
09XDTC1-32@1 5.43 0.47 2.76 0.03 96.5 0.65 0.64 Average 18.76 0.23
09XDTC1-32@2 −7.18 0.40 3.80 0.04
Ol, olivine. Chr, chromite.
09XDTC1-32@3 3.77 0.49 3.86 0.04 a se = standard deviation.
09XDTC1-32@4 2.31 0.41 3.25 0.03 b Data are from Su et al. (2012).
09XDTC1-32@5 −4.17 0.37 3.45 0.03
Average 0.03 3.42
09XD-1@1 4.39 0.30 7.18 0.07 95.2 0.57 0.81
09XD-1@2 12.31 0.42 2.26 0.02
09XD-1@3 13.57 0.43 3.60 0.04
09XD-1@4 6.58 0.34 5.04 0.05
09XD-1@5 12.05 0.44 2.22 0.02
Average 9.78 4.06
09XDTC1-11@1 1.83 0.39 6.39 0.05 92.6 0.76 0.62
09XDTC1-11@2 3.51 0.37 5.85 0.05
09XDTC1-11@3 1.47 0.43 5.57 0.02
09XDTC1-11@4 4.90 0.38 4.25 0.04
Average 2.93 5.52
09XDTC1-31@1 1.65 0.49 6.76 0.05 94.2
09XDTC1-31@2 0.71 0.39 9.46 0.08
09XDTC1-31@3 2.29 0.40 9.82 0.09
09XDTC1-31@4 0.99 0.41 5.94 0.05
Average 1.41 8.00
09XDTC1-19@1 −2.59 0.42 5.95 0.06 93.0 1.18 0.45
09XDTC1-19@2 −6.08 0.39 10.72 0.10
09XDTC1-19@3 2.98 0.58 2.45 0.02
09XDTC1-19@4 −4.80 0.37 11.18 0.14
Average −2.62 7.57
09XDTC1-28@1 11.25 0.48 2.11 0.02 95.7 0.48 1.02
09XDTC1-28@2 17.96 0.81 1.12 0.01
09XDTC1-28@3 11.15 0.52 2.30 0.02
09XDTC1-28@4 11.90 0.53 1.81 0.02
09XDTC1-28@5 12.32 0.56 1.64 0.01
09XDTC1-28@6 14.83 0.75 0.94 0.01
09XDTC1-28@7 23.21 1.37 0.28 0.00
09XDTC1-28@8 16.12 1.15 0.40 0.01
09XDTC1-28@9 18.68 1.84 0.22 0.00
09XDTC1-28@10 20.10 1.15 0.40 0.00
09XDTC1-28@11 19.13 1.08 0.49 0.01
09XDTC1-28@12 20.89 1.03 0.54 0.01
Average 16.46 1.02
09XDTC1-25@1 13.19 1.52 0.24 0.00 96.4 0.22 0.96
09XDTC1-25@2 6.17 1.23 0.37 0.00
09XDTC1-25@3 11.45 0.97 0.60 0.01
09XDTC1-25@4 −0.77 1.18 0.41 0.00
09XDTC1-25@5 16.99 1.85 0.16 0.00
09XDTC1-25@6 11.46 1.70 0.19 0.00
09XDTC1-25@7 12.25 1.80 0.17 0.00
09XDTC1-25@8 13.39 1.17 0.41 0.00
09XDTC1-25@9 11.25 1.52 0.24 0.00
09XDTC1-25@10 7.35 1.87 0.16 0.00
09XDTC1-25@11 7.22 1.26 0.36 0.00
09XDTC1-25@12 −2.88 1.47 0.26 0.00by SIMS. On the contrary, the corresponding Li concentration is appar-
ently higher than the values measured by SIMS, possibly due to the
presence of minor serpentine in the fractures of the olivine separates
(Decitre et al., 2002; Vils et al., 2009). The olivine grains in sample
09XDTC1-35 have the lowest Li concentration and the highest δ7Li
value (Table 1).
4. Discussion
4.1. Primary features of Li isotopes
Lithium isotopes are sensitive to low-temperature fluid–rock inter-
action due to high mobility of Li and large mass difference between 6Li
and 7Li (Chan et al., 2002). During serpentinization, Li and preferentially
6Li are leached from olivine grains to hydrothermal fluid, causing the ol-
ivine to become Li-depleted and isotopically heavier (Chan et al., 2002;
Decitre et al., 2002; Lundstrom et al., 2005; Wimpenny et al., 2010;
Wunder et al., 2010). This process can explain the extreme δ7Li in the ol-
ivine from the highly serpentinized sample 09XDTC1-35, and its subse-
quent shift from the compositional trend defined by themajority of the
samples (Fig. 3). Crustal contamination commonly occurs during40 MC-ICP-MS
09XDTC1-24
09XDTC1-25
30 Arc lavas
SIMS
09XDTC1-16
20 Arc lavas 09XDTC1-15
09XDTC1-32
09XD-1
09XDTC1-11
10 09XDTC1-31
09XDTC1-19
09XDTC1-28
0 09XDTC1-25
09XDTC1-35
09XDTC1-36
-10 09XDTC1-29
0.1 1 10 100
Li (ppm)
Fig. 2. Li versus δ7Li in olivine from the Xiadong dunites. Data of arc lavas plotted for
comparison are from Košler et al. (2009), Chan et al. (2002), Magna et al. (2006), Tang
et al. (2014), and Tomascak et al. (2000, 2002).
7Li (‰)
366 B.-X. Su et al. / Lithos 286–287 (2017) 363–368magma ascent and emplacement, and could potentially elevate the Li
content of magmas and modify their Li isotopic composition (Teng
et al., 2004). For the Xiadong dunite, high MgO and low rare earth ele-
ment concentrations (Su et al., 2014), along with low Li concentrations
of the olivine (Table 1), suggest negligible crustal contamination.
Additionally, chemical exchange between olivine and chromite
plays an important role in modifying their respective original composi-
tions (Bai et al., 2017; Dauphas et al., 2010; Jackson, 1969; Roeder et al.,
1979; Teng et al., 2011; Xiao et al., 2016). It is generally accepted that,
with decreasing temperature, diffusion of Fe occurs from olivine to
chromite, and vice versa for Mg, causing the Fo content of olivine to in-
crease (Bai et al., 2017; Su et al., 2012). Although Lenaz et al. (2017)
showed that the content of Li in spinel from African mantle xenoliths
falls within the range of 0.1–1.5 ppm, our measurement demonstrates
much lower 7Li+ count rates for chromite (several thousand cps/nA)
than those for olivine (hundreds of thousands cps/nA), reflecting the
extremely low Li concentration in the chromite. As a consequence, the
chemical exchange should not affect the olivine in termsof Li concentra-
tion and its isotopic composition. Previous studies (e.g., Lundstrom
et al., 2005; Tomascak et al., 2016) have revealed that diffusion of Li
into a mineral should produce a negative anomaly of around 10–20‰
for δ7Li since 6Li diffuses slightly faster than 7Li. Such diffusion mecha-
nism may account for the negative correlation between Li and δ7Li in
the Xiadong olivine (Fig. 2) but cannot explain variability of the entire
dataset (30‰ variation in δ7Li) and their correlations with indicators
of magma differentiation (Fig. 3).
During the growth of a mineral from a fluid or melt, intra-crystalline
diffusion driven by a chemical or isotopic gradient could affect the Li
concentration of the mineral (Tomascak et al., 2016). During olivine
growth, the establishments of Li concentration and Li isotopic gradients
cause such diffusion to result in uniform Li concentration relative to δ7Li
of olivine in an individual sample, since it takes longer time for Li iso-
topes, compared to Li abundance to achieve compositional uniformity
(Richter et al., 2014). This appears to be consistent with our data,
pointing to the limited Li variations compared to the wide range of
δ7Li values in individual samples (Fig. 2; Table 1). For example, the ho-
mogeneous Fo content relative to Li and δ7Li in a given Xiadong sample
(Su et al., 2012, 2015) can be ascribed to 4–8 times faster Fe-Mga b
10 10
1 1
0.1 09XDTC1-35 0.1 09XDTC1-35
92 93 94 95 96 97 0.4 0.6
Fo in Ol NiO (wt
d 40 e 40
09XDTC1-35 09XDTC1-35
30 30
20 20
10 10
0 0
-10 -10
92 93 94 95 96 97 0.4 0.6
Fo in Ol NiO (wt
Fig. 3. Correlation diagrams of Li and δ7Li vs. Fo in olivine, NiO and MnO in chrom
Li (ppm)
Li (ppm)exchange in olivine than Li diffusion in olivine (Parkinson et al., 2007).
However, the large variations in Fo, Li and δ7Li displayed by our entire
dataset cannot be explained by the intra-crystalline diffusion mecha-
nism (Fig. 3).
Therefore, we conclude that all the olivine grains in the Xiadong du-
nites, except those in sample 09XDTC1-35, preserve the primary Li con-
centration and isotopic compositions during crystallization. Given that
themineral assemblage in the dunites consists of only olivine and chro-
mite, and the fact that the chromite contains little or no Li, we assume
that the Li concentrations and isotopic values in olivine of the Xiadong
dunites are representative of the whole rock samples.4.2. Li isotope fractionation during magma differentiation
Lithium is moderately incompatible during mantle melting and
magma differentiation (Brenan et al., 1998). This behavior indicates
that the concentration of Li inmagmas increasewith increasing degrees
of differentiation (e.g., Hamelin et al., 2009; Marks et al., 2007). This is
consistent with the study of Weyer and Seitz (2012) who analyzed ma-
trix and olivine phenocryst of several basaltic rocks, showing lower Li
concentration in olivine than the matrix in all instances. Large extents
of fractional crystallization of olivine have also been interpreted to ac-
count for the elevated Li contents of some phonolites (Ackerman et al.,
2015). The Xiadong olivine shows a trend of increasing Li with decreas-
ing Fo (Fig. 3a). Furthermore, the Li content of olivine is correlated neg-
atively with NiO and positively with MnO in chromite (Fig. 3b, c).
Collectively, these trends are consistent with variable degrees of
magma differentiation. The results demonstrate that olivine fraction-
ation drives the derivative magma towards high Li. Such elevated Li
trends are thought to be related to redox conditions of the magmas
(Marks et al., 2007). This is evidenced by increasing partition coefficient
of Li between olivine and melt with the increasing concentration of tri-
valent cations in the melt, which are capable of being associated with Li
within the olivine structure (Grant and Wood, 2010). Thus, the highly
oxidized nature of the parentalmagmas of the XiadongAlaskan-type in-
trusion (Su et al., 2012, 2014)may enhance the fractionation of Li in the
olivine.c
10
1
0.1 09XDTC1-35
0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2
.%) in Chr MnO (wt.%) in Chr
f 40
09XDTC1-35
30
20
10
0
-10
0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2
.%) in Chr MnO (wt.%) in Chr
ite for the Xiadong dunites. Arrows represent magma differentiation trend.
Li (ppm)
B.-X. Su et al. / Lithos 286–287 (2017) 363–368 367Theoretically, Li isotope fractionation is produced by different coordi-
nation numbers for Li between melts and crystals (Teng et al., 2006) as
7Li appears to partition favorably into less coordinated sites with stron-
ger Li\\O bonds (Magna et al., 2013; Wunder et al., 2011). Since Li in
most silicatemelts is predominantly in tetrahedrally coordinated groups
(Soltay and Henderson, 2005), minerals incorporating Li in sites of
higher coordination should be isotopically lighter than the melt from
which they crystallize (Teng et al., 2006). For example, granitic melts
should evolve towards high δ7Li values with differentiation because Li
is octahedrally coordinated in most Li-rich minerals (Teng et al., 2006).
However, a compilation of global igneous rocks has revealed nega-
tive correlations between δ7Li and Li (Tomascak et al., 2016), indicating
that Li isotopic compositions tend to become lighter during magma
differentiation (Magna et al., 2016). In addition, more geochemically
differentiated samples of a suite of evolved granites with the lightest
Li isotopic signatures revealed Li isotope fractionation during magma
differentiation (Plyusnin et al., 1979). A study of differentiated litholo-
gies of the Massif Central in France also revealed a negative correlation
between Li isotopic composition (δ7Li ≥ +0.4‰) and Li concentration,
an observation consistent with fractional crystallization (Hamelin
et al., 2009).
Because of similar sizes of Li and Mg cations (De Hoog et al., 2010),
Li+ preferentially substitutes for Mg2+ in six-fold coordinated site in ol-
ivine (Grant andWood, 2010). Thus, olivine is expected to be isotopical-
ly lighter than the basaltic matrix, as Li in basaltic melts is believed to be
in tetrahedrally coordinated groups (Soltay and Henderson, 2005). Es-
sentially, the olivine phenocrysts analyzed by Weyer and Seitz (2012)
show larger δ7Li variation (−10.5 to +6.5‰) than the matrix (+2.5
and +4.7‰). The lower δ7Li values of the olivine may have been pro-
duced by diffusion process from high-Li matrix to olivine, whereas the
higher δ7Li values probably represent the initial compositions of the ol-
ivine phenocrysts after crystallization. Becausemantle-derivedmagmas
reported so far have lower δ7Li values (summary in Tomascak et al.,
2016) than the olivine in this study, the much higher δ7Li signature
in the Xiadong olivine (Fig. 2) cannot be explained by theoretical
calculation. Diffusion alone is not able to account for the observed Li iso-
topic compositions. Consequently, these olivine grains preserve primary
magmatic compositionswith respect to Li isotopes. The trends shown in
Fig. 3d, e and f illustrate that magma differentiation caused the crystal-
lized olivine as well as the differentiating magmas to become lighter in
Li isotopes.
According to Tomascak et al. (2016), the discrepancies between the-
oretical calculation and actual observation through measured results
could be explained by the limited application of the theoretical
approach to natural open-system and diffusion effect. Additionally,
Teng et al. (2006) highlighted that fluid exsolution could be responsible
for the significant Li isotope fractionation in granitic systems. Apparent-
ly, the magma system of the Xiadong Alaskan-type intrusion is a quasi-
closed, hydrous system (Su et al., 2012), which possibly reconciles the
discrepancies.
4.3. Implications for arc magmatism
It has been well documented that Li behaves as a fluid-mobile ele-
ment in aqueous transfer (Brenan et al., 1998) with 7Li preferentially
enriched in the fluid phase (e.g., Chan et al., 1999; Xiao et al., 2015).
Therefore, the hydrous reservoirs are generally isotopically heavy
(Tomascak et al., 2016). The inference is manifested in the observations
of the Xiadong intrusion. Represented by olivine chemistry, the whole-
rock δ7Li values of the dunites from the Xiadong Alaskan-type intrusion
are apparently higher than those in mantle peridotites and volcanic
rocks and are comparable to surface water and marine sediments
(Tomascak et al., 2016). The Li isotopic signature of the Xiadong dunites
is compatiblewith the hydrous nature of the parentalmagmas of the in-
trusion, originating from a mantle wedge metasomatized by slab-
derived fluids (Su et al., 2012, 2014).Arc magmas have Li and δ7Li ranging from 1 to 50 ppm and 0 to
+8‰, respectively (e.g., Magna et al., 2006; Tang et al., 2014;
Tomascak et al., 2000, 2002), overlapping the ranges of themost evolved
olivine in theXiadongdunites (Fig. 2). This similarity indicates that Li iso-
topic compositions of arc magmas are likely related to fractional crystal-
lization of olivine. The cumulus olivine with heavy Li isotopes are most
likely stored in the deep arc crust. Lithium isotope fractionation during
arc magma differentiation is probably favored by high oxygen fugacity
(Marks et al., 2007), high fluid enrichment (Teng et al., 2006), and highly
differentiated lithology (Hamelin et al., 2009; Plyusnin et al., 1979), all of
which are typical of the Alaskan-type complexes (Farahat and Helmy,
2006; Himmelberg and Loney, 1995; Irvine, 1974; Su et al., 2014). The
above-mentioned conditions and characteristics are consistent with the
geochemical behavior of Li isotopes in subduction zone, which further
confirms the findings of previous studies (e.g., Marschall et al., 2007;
Wunder et al., 2006; Xiao et al., 2015; Yamaji et al., 2001) regarding
the preferential loss of isotopically heavy Li from subducted slab to the
mantle wedge during progressive dehydration, leaving behind isotopi-
cally lighter residual slab.
5. Conclusions
The Xiadong intrusion in NW China is a well characterized Alaskan-
type intrusion and represents cumulates in the early differentiation of
mafic-ultramafic magmas. Lithium isotope analyses of olivine from 13
dunites of the intrusion yield the following conclusions:
(1) Olivine grains in the dunites have Li and δ7Li variations ranging
from 0.10 to 11.18 ppm and −7.18 to +34.4‰, respectively.
The inter-sample and intra-crystal variations cannot be ascribed
to alteration, crustal contamination, compositional exchange, or
diffusion processes, but mainly represent primary Li concentra-
tions and isotopic compositions during crystallization.
(2) The Li concentrations and δ7Li values are well correlated with
mineral chemical indices of magma differentiation, indicating Li
isotope fractionation duringmagma differentiation. The decreas-
ing δ7Li value with increase of Li concentration is in accord with
magma differentiation.
(3) The compositional overlap between arc magmas and the most
evolved olivine in the Xiadong dunites suggests that primitive
arc magmas might have been initially heavier in Li isotopes and
later underwent removal of isotopically heavy olivine at depth.
Acknowledgements
We appreciate Paul Tomascak, Tomas Magna, Davide Lenaz, Celine
Martin and one anonymous reviewer for their constructive reviews
and scientific inputs on the early version of the manuscript, Nathalie
Vigier and Etienne Deloule for Li isotope analysis using MC-ICP-MS
in CRPG. This work was financially supported by the National Natural
Science Foundation of China (41522203 and 41173011), National Key
Research and Development Program of China (2017YFC0601204) and
Youth Innovation Promotion Association, Chinese Academy of Sciences
(2016067).
References
Ackerman, L., Ulrych, J., Řanda, Z., Erban, V., Hegner, E., Magna, T., Balogh, K., Frána, J.,
Lang, M., Novák, J.K., 2015. Geochemical characteristics and petrogenesis of phono-
lites and trachytic rocks from the České Středohoří Volcanic Complex, the Ohře Rift,
Bohemian Massif. Lithos 224–225, 256–271.
Bai, Y., Su, B.X., Chen, C., Yang, S.H., Liang, Z., Xiao, Y., Qin, K.Z., Malaviarachchi, S.P.K., 2017.
Base metal mineral segregation and Fe-Mg exchange inducing extreme compositions
of olivine and chromite from the Xiadong Alaskan-type complex in the southern part
of the Central Asian Orogenic Belt. Ore Geology Reviews. http://dx.doi.org/10.1016/j.
oregeorev.2017.01.023.
Brenan, J.M., Neroda, E., Lundstrom, C.C., Shaw, H.F., Ryerson, F.J., Phinney, D.L., 1998.
Behaviour of boron, beryllium and lithium during melting and crystallization:
368 B.-X. Su et al. / Lithos 286–287 (2017) 363–368constraints from mineral-melt partitioning experiments. Geochimica et
Cosmochimica Acta 62, 2129–2141.
Chacko, T., Cole, D.R., Horita, J., 2001. In: Valley, J.W., Cole, D.R. (Eds.), EquilibriumOxygen,
Hydrogen and Carbon Isotope Fractionation Factors Applicable to Geological Systems.
Stable Isotope Geochemistry, Reviews in Mineralogy and Geochemistry. vol. 43.
Mineralogical Society of America, Washington, DC, pp. 1–82.
Chan, L.H., Frey, F.A., 2003. Lithium isotope geochemistry of the Hawaiian plume: results
from the Hawaii Scientific Drilling Project and Koolau volcano. Geochemistry,
Geophysics, Geosystems 4.
Chan, L.H., Leeman, W.P., You, C.F., 1999. Lithium isotopic composition of Central
American Volcanic Arc lavas: implications for modification of subarc mantle by
slab-derived fluids. Chemical Geology 160, 255–280.
Chan, L.H., Leeman, W.P., You, C.F., 2002. Lithium isotopic composition of Central
American volcanic arc lavas: implications for modification of subarc mantle by slab-
derived fluids: correction. Chemical Geology 182, 293–300.
Dauphas, N., Teng, F.Z., Arndt, N.T., 2010. Magnesium and iron isotopes in 2.7 Ga Alexo
komatiites: mantle signatures, no evidence for Soret diffusion, and identification of dif-
fusive transport in zoned olivine. Geochimica et Cosmochimica Acta 74, 3274–3291.
Day, J., Qiu, L., Ash, R.D., McDonough, W.F., Teng, F.Z., Rudnick, R.L., Taylor, L.A., 2016.
Evidence for high-temperature fractionation of lithium isotopes during differentia-
tion of the moon. Meteoritics & Planetary Science 51, 1046–1062.
De Hoog, J.C., Gall, L., Cornell, D.H., 2010. Trace-element geochemistry of mantle olivine
and application to mantle petrogenesis and geothermobarometry. Chemical Geology
270, 196–215.
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. Geochemistry,
Geophysics, Geosystems 3. http://dx.doi.org/10.1029/2001GC000178.
Farahat, E.S., Helmy, H.M., 2006. Abu Hamamid Neoproterozoic Alaskan-type complex,
south Eastern Desert, Egypt. Journal of African Earth Sciences 45, 187–197.
Grant, K.J., Wood, B.J., 2010. Experimental study of the incorporation of Li, Sc, Al and other
trace elements into olivine. Geochimica et Cosmochimica Acta 74, 2412–2428.
Hamelin, C., Seitz, H.M., Barrat, J.A., Dosso, L., Maury, R., Chaussidon, M., 2009. A low δ7Li
lower crustal component: evidence from an alkalic intraplate volcanic series (Chaîne
des Puys, French Massif Central). Chemical Geology 266, 205–217.
Himmelberg, G.R., Loney, R.A., 1995. Characteristics and Petrogenesis of Alaskan-type
Ultramafic-mafic Intrusions, Southeastern Alaska. US Geological Survey, Professional
Papers 1564 pp. 1–47.
Irvine, T.N., 1974. Petrology of the Duke Island ultramafic complex southern Alaska.
Geological Society of America Memoirs 138, 240.
Jackson, E.D., 1969. Chemical variation in coexisting chromite and olivine in chromitite
zones of Stillwater complex. Economic Geology Monographs 4, 41–71.
Košler, J., Magna, T., Mlčoch, B., Mixa, P., Nývlt, D., Holub, F.V., 2009. Combined Sr, Nd, Pb
and Li isotope geochemistry of alkaline lavas from northern James Ross Island
(Antarctic Peninsula) and implications for back-arc magma formation. Chemical
Geology 258, 207–218.
Lenaz, D., Musco, M.E., Petrelli, M., Caldeira, R., DeMin, A., Marzoli, A., Mata, J., Perugini, D.,
Princivalle, F., Boumehdi, M.A., Bensaid, I.A.A., Youbi, N., 2017. Restitic or not? Insights
from trace element content and crystal—structure of spinels in African mantle xeno-
liths. Lithos 278, 464–476.
Lundstrom, C.C., Chaussidon, M., Hsui, A.T., Kelemen, P., Zimmerman, M., 2005. Observa-
tions of Li isotopic variations in the Trinity Ophiolite: evidence for isotopic fraction-
ation by diffusion during mantle melting. Geochimica et Cosmochimica Acta 69,
735–751.
Magna, T., Wiechert, U., Grove, T.L., Halliday, A.N., 2006. Lithium isotope fractionation in
the southern Cascadia subduction zone. Earth and Planetary Science Letters 250,
428–443.
Magna, T., Novák, M., Janoušek, V., 2013. Lithium isotopes in giant pegmatite
bodies—implications for their sources and evolution. Geological Association of
Canada and Mineralogical Association of Canada Annual Meeting, Winnipeg,
Canada. Abstract Volume, p. 135.
Magna, T., Novák, M., Cempírek, J., Janoušek, V., Ullmann, C.V., Wiechert, U., 2016.
Crystallographic control on lithium isotope fractionation in Archean to Cenozoic
lithium-cesium-tantalum pegmatites. Geology 44, 655–658.
Marks, M.A.W., Rudnick, R.L., McCammon, C., Vennemann, T., Markl, G., 2007. Arrested
kinetic Li isotope fractionation at the margin of the Ilímaussaq complex, South
Greenland: evidence for open-system processes during final cooling of peralkaline ig-
neous rocks. Chemical Geology 246, 207–230.
Marschall, H.R., Pogge von Strandmann, P.A.E., Seitz, H.M., Elliott, T., Niu, Y.L., 2007. The
lithium isotopic composition of orogenic eclogites and deep subducted slabs. Earth
and Planetary Science Letters 262, 563–580.
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.
Plyusnin, G.S., Posokhov, V.F., Sandimirova, G.P., 1979. Magmatic differentiation and rela-
tionship of 7Li/6Li ratio to fluorine content. Doklady Academy of Sciences USSR, Earth
Sciences 248, 187–189.
Richter, F., Watson, B., Chaussidon, M., Mendybaev, R., Ruscitto, D., 2014. Lithium isotope
fractionation by diffusion in minerals. Part 1: pyroxenes. Geochimica et
Cosmochimica Acta 126, 352–370.
Roeder, P.L., Campbell, I.H., Jamieson, H.E., 1979. Re-evaluation of the olivine-spinel
geothermometer. Contributions to Mineralogy and Petrology 68, 325–334.Schuessler, J.A., Schoenberg, R., Sigmarsson, O., 2009. Iron and lithium isotope systematics
of the Hekla volcano, Iceland—evidence for Fe isotope fractionation during magma
differentiation. Chemical Geology 258, 78–91.
Soltay, L.G., Henderson, G.S., 2005. Structural differences between lithium silicate and
lithium germanate glasses by Raman spectroscopy. Physics and Chemistry of Glasses
46, 381–384.
Su, B.X., Qin, K.Z., Sakyi, P.A., Malaviarachchi, S.P.K., Liu, P.P., Tang, D.M., Xiao, Q.H., Sun, H.,
Ma, Y.G., Mao, Q., 2012. Occurrence of an Alaskan-type complex in the Middle
Tianshan Massif, Central Asian Orogenic Belt: inferences from petrological and min-
eralogical studies. International Geology Review 54, 249–269.
Su, B.X., Qin, K.Z., Zhou, M.F., Sakyi, P.A., Thakurta, J., Tang, D.M., Liu, P.P., Xiao, Q.H.,
Sun, H., 2014. Petrological, geochemical and geochronological constraints on
the origin of the Xiadong Ural-Alaskan type complex in NWChina and tectonic impli-
cation for the evolution of southern Central Asian Orogenic Belt. Lithos 200–201,
226–240.
Su, B.X., Gu, X.Y., Deloule, E., Zhang, H.F., Li, Q.L., Li, X.H., Vigier, N., Tang, Y.J., Tang, G.Q.,
Liu, Y., Brewer, A., Mao, Q., Ma, Y.G., 2015. Potential orthopyroxene, clinopyroxene
and olivine referencematerials for in situ lithium isotope determination. Geostandards
and Geoanalytical Research 39, 357–369.
Su, B.X., Zhou, M.F., Robinson, P.T., 2016. Extremely large fractionation of Li isotopes in
chromitite-bearing mantle sequence. Scientific Reports 6, 22370.
Sun, H., Qin, K.Z., Su, B.X., Fan, X., Tang, D.M., Li, J., 2009. Discovery of komatiitic ultramafic
intrusion in Mid-Tianshan terrain: Xiadong intrusion, Xinjiang. Acta Petrologica
Sinica 25, 738–748 (in Chinese with English abstract).
Tang, M., Rudnick, R.L., Chauvel, C., 2014. Sedimentary input to the source of Lesser
Antilles lavas: a Li perspective. Geochimica et Cosmochimica Acta 144, 43–58.
Taylor, H.P., 1967. The zoned ultramafic complexes of southeastern Alaska. In: Wyllie, P.J.
(Ed.), Ultramafic and Related Rocks. Wiley, New York, pp. 97–121.
Teng, F.Z., McDonough, W.F., Rudnick, R.L., Dalpé, C., Tomascak, P.B., Chappell, B.W., Gao,
S., 2004. Lithium isotopic composition and concentration of the upper continental
crust. Geochimica et Cosmochimica Acta 68, 4167–4178.
Teng, F.Z., McDonough, W.F., Rudnick, R.L., Walker, R.J., 2006. Diffusion-driven extreme
lithium isotopic fractionation in country rocks of the Tin Mountain pegmatite. Earth
and Planetary Science Letters 243, 701–710.
Teng, F.Z., Dauphas, N., Helz, R.T., Gao, S., Huang, S., 2011. Diffusion-driven magnesium
and iron isotope fractionation in Hawaiian olivine. Earth and Planetary Science
Letters 308, 317–324.
Tomascak, P.B., Tera, F., Helz, R.T., Walker, R.J., 1999. The absence of lithium isotope frac-
tionation during basalt differentiation: new measurements by multi-collector sector
ICP-MS. Geochimica et Cosmochimica Acta 63, 907–910.
Tomascak, P.B., Ryan, J.G., Defant, M.J., 2000. Lithium isotope evidence for light element
decoupling in the Panama subarc mantle. Geology 28, 507–510.
Tomascak, P.B., Widom, E., Benton, L.D., Goldstein, S.L., Ryan, J.G., 2002. The control of lith-
ium budgets in island arcs. Earth and Planetary Science Letters 196, 227–238.
Tomascak, P.B., Magna, T., Dohmen, R., 2016. Advances in Lithium Isotope Geochemistry.
Springer International Publishing Switzerland, pp. 1–195.
Vigier, N., Decarreau, A., Millot, R., Carignan, J., Petit, S., France-Lanord, C., 2008.
Quantifying Li isotope fractionation during smectite formation and implications for
the Li cycle. Geochimica et Cosmochimica Acta 72, 780–792.
Vils, F., Tonarini, S., Kalt, A., Seitz, H.M., 2009. Boron, lithium and strontium isotopes as
tracers of seawater–serpentinite interaction at Mid-Atlantic ridge, ODP Leg 209.
Earth and Planetary Science Letters 286, 414–425.
Weyer, S., Seitz, H.M., 2012. Coupled lithium- and iron isotope fractionation during mag-
matic differentiation. Chemical Geology 294–295, 42–50.
White, W.M., 2015. Isotope Geochemistry. John Wiley and Sons.
Wimpenny, J., Gislason, S.R., James, R.H., Gannoun, A., Pogge von Strandmann, P.A.E.,
Burton, K., 2010. The behaviour of Li and Mg isotopes during primary phase dissolu-
tion and secondary mineral formation in basalt. Geochimica et Cosmochimica Acta
74, 5259–5279.
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.
Wunder, B., Deschamps, F., Watenphul, A., Guillot, S., Meixner, A., Romer, R.L., Wirth, R.,
2010. The effect of chrysotile nano-tubes on the serpentine-fluid Li-isotopic fraction-
ation. Contributions to Mineralogy and Petrology 159, 781–790.
Wunder, B., Romer, R.L., Meixner, A., Jahn, S., 2011. Li-isotope silicate fluid fractionation:
pressure dependence and influence of the bonding environment. European Journal
of Mineralogy 23, 333–342.
Xiao, Y.L., Sun, H., Gu, H.O., Huang, J., Li, W.Y., Liu, L., 2015. Fluid/melt in continental deep
subduction zones: compositions and related geochemical fractionations. Science
China: Earth Sciences 58, 1457–1476.
Xiao, Y., Teng, F.Z., Su, B.X., Hu, Y., Zhou, M.F., Zhu, B., Shi, R.D., Huang, Q.S., Gong, X.H., He,
Y.S., 2016. Iron and magnesium isotopic constraints on the origin of chemical hetero-
geneity in podiform chromitite from the Luobusa ophiolite, Tibet. Geochemistry, Geo-
physics, Geosystems 17, 940–953.
Yamaji, K., Makita, Y., Watanabe, H., Sonoda, A., Kanoh, H., Hirotsu, T., Ooi, K., 2001. The-
oretical estimation of lithium isotopic reduced partition function ratio for lithium
ions in aqueous solution. Journal of Physical and Chemical Annual 105, 602–613.