Magnesium isotope systematics in Martian meteorites

JID:EPSL

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.1 (1-8)

Earth and Planetary Science Letters ••• (••••) •••–•••

Contents lists available at ScienceDirect

Earth and Planetary Science Letters www.elsevier.com/locate/epsl

Magnesium isotope systematics in Martian meteorites Tomáš Magna a,b,∗ , Yan Hu c , Fang-Zhen Teng c , Klaus Mezger a,d a

Institut für Mineralogie, Universität Münster, Corrensstr. 24, D-48149 Münster, Germany Czech Geological Survey, Klárov 3, CZ-118 21 Prague 1, Czech Republic c Isotope Laboratory, Department of Earth and Space Sciences, University of Washington, Seattle, WA 98195, USA d Institut für Geologie, Universität Bern, Baltzerstr. 1+3, CH-3012 Bern, Switzerland b

a r t i c l e

i n f o

Article history: Received 23 November 2016 Received in revised form 3 July 2017 Accepted 5 July 2017 Available online xxxx Editor: F. Moynier Keywords: magnesium isotopes Mars mantle crust surface

a b s t r a c t Magnesium isotope compositions are reported for a suite of Martian meteorites that span the range of petrological and geochemical types recognized to date for Mars, including crustal breccia Northwest Africa (NWA) 7034. The δ 26 Mg values (per mil units relative to DSM-3 reference material) range from −0.32 to −0.11h; basaltic shergottites and nakhlites lie to the heavier end of the Mg isotope range whereas olivine-phyric, olivine–orthopyroxene-phyric and lherzolitic shergottites, and chassignites have slightly lighter Mg isotope compositions, attesting to modest correlation of Mg isotopes and petrology of the samples. Slightly heavier Mg isotope compositions found for surface-related materials (NWA 7034, black glass fraction of the Tissint shergottite fall; δ 26 Mg > −0.17h) indicate measurable Mg isotope difference between the Martian mantle and crust but the true extent of Mg isotope fractionation for Martian surface materials remains unconstrained. The range of δ 26 Mg values from −0.19 to −0.11h in nakhlites is most likely due to accumulation of clinopyroxene during petrogenesis rather than garnet fractionation in the source or assimilation of surface material modified at low temperatures. The rather restricted range in Mg isotope compositions between spatially and temporally distinct mantle-derived samples supports the idea of inefficient/absent major tectonic cycles on Mars, which would include plate tectonics and large-scale recycling of isotopically fractionated surface materials back into the Martian mantle. The cumulative δ 26 Mg value of Martian samples, which are not influenced by late-stage alteration processes and/or crust–mantle interactions, is −0.271 ± 0.040h (2SD) and is considered to reflect δ 26 Mg value of the Bulk Silicate Mars. This value is robust taking into account the range of lithologies involved in this estimate. It also attests to the lack of the Mg isotope variability reported for the inner Solar System bodies at current analytical precision, also noted for several other major elements. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Together with Si, O and Fe, Mg belongs to the most abundant elements in the rocky planets of the inner Solar System (e.g., Palme and O’Neill, 2014). Isotope studies on these elements reveal different degrees of mass-dependent fractionation that are the result of a variety of geochemical and cosmochemical processes. The Mg isotope systematics of major silicate reservoirs in the Solar System have been discussed in a number of studies that have shown the absence of detectable Mg isotope fractionation during planet formation and identical Mg isotope composition for the Moon, chondrites, and achondrites (e.g., Bourdon et al., 2010; Chakrabarti and Jacobsen, 2010; Sedaghatpour et al., 2013; Teng et al., 2010)

*

Corresponding author at: Czech Geological Survey, Klárov 3, CZ-118 21 Prague 1, Czech Republic. E-mail address: [email protected] (T. Magna). http://dx.doi.org/10.1016/j.epsl.2017.07.012 0012-821X/© 2017 Elsevier B.V. All rights reserved.

which are all similar to the estimated δ 26 Mg = −0.25 ± 0.07h for the Earth’s mantle (Teng et al., 2010). This isotope homogeneity in large planetary bodies has also been observed for Ca (Magna et al., 2015b; Valdes et al., 2014) while the stable isotope compositions of some other major elements (Si, Fe) display slight variations (e.g., Dauphas et al., 2015; Pringle et al., 2014; Sossi et al., 2016), possibly as a consequence of volatile loss and/or core formation. Magnesium isotope fractionation is limited during high-temperature processes while low-temperature processes such as alteration, weathering and/or carbonate/sulfate formation may impart sizeable Mg isotope fractionation (see Teng, 2017, and references therein). Magnesium isotope fractionation has been observed during large-scale silicate differentiation (Sedaghatpour et al., 2013; Schiller et al., 2017), but this may also be attributed to the influence of material exposed to low-temperature surface processes (Yang et al., 2016). This contrasting behavior of Mg isotopes in surface versus mantle processes allows for the reconstruction of

JID:EPSL

AID:14554 /SCO

2

[m5G; v1.221; Prn:26/07/2017; 8:07] P.2 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

processes involving chemical exchange between various reservoirs during planetary differentiation. On Earth, surface materials enter the mantle via plate tectonic processes. This can be observed in melt products derived from the mantle that bear evidence for incorporation of surfaceweathered materials. In this respect, Mg isotopes may potentially be employed as a tracer of recycling (Huang et al., 2016; Ke et al., 2016). Therefore, the investigation of Mg isotope systematics of Martian magmatic rocks has the potential both to provide insights into the extent of surface–mantle interaction and to reveal the Mg isotope composition of the bulk Mars. Mars, which has about one-eighth the mass of the Earth, is a unique terrestrial planet that has undergone all planetary evolutionary stages but did not develop plate tectonics (Breuer and Spohn, 2003). Previous studies of Mg isotope composition of Martian meteorites have been inconsistent between individual laboratories (Bizzarro et al., 2011; Chakrabarti and Jacobsen, 2010; Wiechert and Halliday, 2007), which has so far prevented evaluating the Mg isotope composition of Mars. Despite these issues, Mg isotope ratios determined on materials from Mars are broadly identical to the range observed for the Earth, Moon, and chondrites, suggesting a homogeneous Mg isotope distribution within the inner Solar System. This study presents the first comprehensive and systematic survey of Mg isotope compositions for a total of 31 Martian meteorites that span a range of Martian lithologies recognized to date, including materials compositionally similar to Martian crust. Combined with the published Mg isotope data for the Earth, Moon, chondrites and other planetary bodies of the inner Solar System (e.g., Vesta), this study provides further constraints on the Mg isotope (in)variability of mantles of the inner Solar System planetary bodies and on the interaction of surface material with mantle rocks on Mars, which contribute to our understanding of planetary evolution. 2. Samples and methods Magnesium isotope compositions were obtained for 21 shergottites, six clinopyroxene-bearing nakhlites [Nakhla, Lafayette, Miller Range (MIL) 03346, Yamato (Y)-000593, Northwest Africa (NWA) 817, NWA 5790], two dunitic chassignites (Chassigny, NWA 2737), orthopyroxenite Allan Hills (ALH) 84001 and polymict crustal breccia NWA 7034 (Agee et al., 2013). Shergottites include five basaltic (Shergotty, Zagami, Los Angeles, NWA 856, NWA 4864), ten olivine-phyric [Elephant Moraine (EETA) 79001 lithology A, Larkman Nunataks (LAR) 06319, Y-980459, NWA 1068, NWA 4925, NWA 6162, Sayh al Uhaymir (SaU) 005, SaU 051, SaU 094, Tissint], two olivine–orthopyroxene-phyric [Roberts Massif (RBT) 04262, Dar al Gani (DaG) 476], three lherzolitic (ALH 77005, Y-000097, NWA 1950), and one diabasic shergottite (NWA 5990). Fusion crusts were removed from samples and only interior portions of the meteorites were processed. These samples were previously characterized for their Li and Ca isotope systematics (Magna et al. 2015a, 2015b). Magnesium isotope compositions of Martian meteorites were measured at the University of Arkansas, with the exception of Tissint and NWA 7034, which were analyzed at the University of Washington. The analytical procedures for the chemical separation and isotope analysis of Mg followed those reported elsewhere (Teng et al. 2010, 2015). Aliquots of dissolved Martian meteorites from the previous study of Magna et al. (2015a) were used. After drying down, residues were refluxed with concentrated HNO3 and then diluted to 1 M HNO3 solution for cation exchange chromatography. Pure Mg solutions were obtained by two passes through quartz glass columns packed with Bio-Rad AG50W-X8 resin (200–400 mesh) in 1 M HNO3 . Total procedure blank (<10 ng) was considered negligible com-

Fig. 1. Magnesium three-isotope plot for the Martian meteorite suite from this study. The dashed line represents the fractionation line with a slope of 0.515. The Mg isotope composition of the Earth’s mantle is from Teng et al. (2010).

pared to Mg processed. At least two reference materials were processed together with unknown samples for each batch of column chemistry. Magnesium isotope compositions were measured using the standard–sample bracketing method on a Nu Plasma multiple-collector inductively-coupled-plasma mass spectrometer (MC-ICPMS), housed at the University of Arkansas, and a Nu Plasma II MC-ICPMS, housed at the University of Washington, respectively. The Mg isotope data are expressed relative to the DSM-3 reference material (Galy et al., 2003) and reported as δ x Mg (h) = [(x Mg/24 Mg)sample /(x Mg/24 Mg)DSM-3 − 1] × 1000, where x Mg denotes 25 Mg and 26 Mg, respectively. Throughout this study, δ 26 Mg value is used considering mass-dependent fractionation within analytical uncertainties. The external reproducibility of Mg isotope measurements was better than ±0.07h (2σ ) for δ 26 Mg and ±0.05h (2σ ) for δ 25 Mg, verified by duplicate and replicate measurements of synthetic solutions over the four-year period (Teng et al., 2015). The Mg isotope compositions for reference materials (Kilbourne Hole olivine: δ 26 Mg = −0.268 ± 0.029h; Hawaiian seawater: δ 26 Mg = −0.836 ± 0.046h; Murchison chondrite: δ 26 Mg = −0.290 ± 0.044h) agree with published values (see Teng et al., 2015, and references therein). 3. Results The Mg isotope compositions of Martian meteorites are listed in Table 1. The δ 25 Mg and δ 26 Mg values range from −0.156 to −0.055h and −0.318 to −0.106h, respectively, and follow the mass-dependent fractionation line (Fig. 1). The new Mg isotope data for Zagami, Nakhla, Chassigny and ALH 84001 are consistently intermediate between those reported by Wiechert and Halliday (2007) and Chakrabarti and Jacobsen (2010) but we note that their data were critically reviewed elsewhere (e.g., Bizzarro et al., 2011; Teng et al., 2015) and considered an analytical artifact. The Mg isotope composition of NWA 856 is identical within the analytical error to that reported by Bizzarro et al. (2011). In contrast to terrestrial peridotites, OIB and MORB (Bourdon et al., 2010; Teng et al., 2010), Martian meteorites display measurable Mg isotope variations, reflecting their petrological diversity. Basaltic shergottites and nakhlites have among the highest δ 26 Mg values of the entire suite whereas olivine-phyric, olivine–

JID:EPSL

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.3 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

3

Table 1 Magnesium isotope compositions of Martian meteorites.

Basaltic shergottites Los Angeles 001 NWA 856

Notes

MgO (wt.%)

δ 26 Mg (h)

2SD

δ 25 Mg (h)

2SD

LaN /SmN

USNM 7058

3.5 9.5

−0.108 −0.193 −0.258 −0.171 −0.205 −0.261 −0.09 −0.13 −0.55

0.065 0.065 0.018 0.065 0.069 0.069 0.12 0.11 0.10

−0.085 −0.123 −0.134 −0.055 −0.096 −0.138 −0.03 −0.08 −0.28

0.048 0.048 0.006 0.048 0.057 0.057 0.12 0.09 0.06

0.94 0.90

0.068 0.055 0.055 0.077 0.077 0.055 0.077 0.077 0.067 0.072 0.070 0.064 0.055

−0.122 −0.125 −0.126 −0.114 −0.146 −0.122 −0.138 −0.156 −0.143 −0.141 −0.082 −0.060 −0.151

0.064 0.056 0.056 0.073 0.073 0.056 0.073 0.073 0.043 0.038 0.046 0.047 0.056

0.34 0.99 0.95 0.17 0.36 0.16 0.16 0.16 0.20

Bizzarro et al. (2011) NWA 4864 Shergotty (F) Zagami (F)

USNM 6676

8.1a 8.7 12.3

Wiechert and Halliday (2007) Wiechert and Halliday (2007) Chakrabarti and Jacobsen (2010) Olivine-phyric shergottites EETA 79001A LAR 06319 NWA 1068 NWA 4925 NWA 6162 SaU 005 SaU 051 SaU 094 Tissint (F)

0.92 0.94

18.7

−0.253 −0.280 −0.222 −0.181 −0.255 −0.267 −0.238 −0.282 −0.292 −0.307 −0.160 −0.121 −0.283

Olivine–orthopyroxene-phyric shergottites DaG 476 RBT 04262

20.8 21.6

−0.260 −0.318

0.072 0.053

−0.111 −0.143

0.054 0.048

0.19 0.89

Diabasic shergottittes NWA 5990

17.1

−0.106

0.058

−0.071

0.044

0.15

28.7 25.1 25.8

−0.249 −0.275 −0.264

0.074 0.055 0.077

−0.130 −0.12 −0.144

0.069 0.056 0.073

0.44 0.50 0.45

12.9 9.3 12.1

−0.195 −0.134 −0.113 −0.07 −0.58 −0.129 −0.155 −0.149

0.072 0.065 0.072 0.03 0.07 0.059 0.065 0.055

−0.121 −0.092 −0.064 0.00 −0.31 −0.056 −0.108 −0.069

0.054 0.048 0.054 0.02 0.04 0.051 0.048 0.056

1.39 1.33 1.68

−0.272 −0.05 −0.57 −0.283

0.072 0.07 0.08 0.061

−0.121 0.00 −0.30 −0.134

0.054 0.04 0.05 0.047

Y-980459

Lherzolitic shergottites ALH 77005 NWA 1950 Y-000097

NASA, subsample 644 NASA, subsample 47

MPI 1522/2

22.7 17.1b replicate black glass-rich fraction

20.9

NIPR, subsample 53

NASA, subsample 223 NIPR, subsample 71

Nakhlites (clinopyroxenites) Lafayette USNM 1505 MIL 03346 NASA, subsample 193 Nakhla (F) Wiechert and Halliday (2007) Chakrabarti and Jacobsen (2010) NWA 817 NWA 5790 Y-000593 NIPR, subsample 121 Chassignites (dunites) Chassigny (F)

NHMV, subsample H 3398 Wiechert and Halliday (2007) Chakrabarti and Jacobsen (2010)

NWA 2737 Orthopyroxenites ALH 84001

Other non-SNC samples NWA 7034

16.1 15.8 16.5 15.9 21.8 20.5

10.3 8.0 10.4

31.8

37.1

1.12 0.22

1.88 2.06 1.60

2.37

2.76

NASA, subsample 413 Wiechert and Halliday (2007)

25.0

−0.189 −0.12

0.077 0.09

−0.059 −0.04

0.073 0.10

0.99

polymict breccia duplicate/replicate

7.8c

−0.170 −0.166

0.071 0.143

−0.087 −0.067

0.053 0.082

1.33

“F” denotes an observed fall. The degree of enrichment/depletion is based on LaN /SmN (>0.7 for enriched, 0.3–0.7 for intermediate and <0.3 for depleted shergottites; see Magna et al., 2015a, for references). 2SD represents two times the standard deviation of the population of repeat measurements from the bracketing pure Mg standards during an analytical session. We refer to Teng et al. (2015) for further details. a Fusion crust of paired shergottite NWA 2795. See discussion in He et al. (2015). b Data from Chennaoui Aoudjehane et al. (2012). c Plumose groundmass of Agee et al. (2013).

orthopyroxene-phyric and lherzolitic shergottites, and chassignites are isotopically lighter (Fig. 1), likely dominated by Mg hosted in olivine. The low δ 26 Mg of −0.195 ± 0.072h has been found for the nakhlite Lafayette (cf. other nakhlites with δ 26 Mg from −0.155 ± 0.065h to −0.113 ± 0.072h; Table 1) which may be associated with a slightly larger proportion of olivine compared with the other nakhlites. The orthopyroxenite ALH 84001 has in-

termediate δ 26 Mg = −0.189 ± 0.077h. The polymict crustal breccia NWA 7034 has δ 26 Mg at the heavier end of the entire range found for Martian meteorites (−0.170 ± 0.071h) and a sub-sample of the Tissint fall with abundant black glass that appears to have experienced some processing on the surface of Mars (see Chennaoui Aoudjehane et al., 2012, for discussion), has δ 26 Mg = −0.160 ± 0.070h, which is identical to that of NWA 7034 (Fig. 1).

JID:EPSL

4

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.4 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

Fig. 2. The δ 26 Mg values versus MgO contents (a), and modal plagioclase (b) of Martian meteorites from this study. The MgO content and δ 26 Mg value of Bulk Silicate Mars (BSM) are from McSween and McLennan (2014) and this study, respectively. The MgO content and δ 26 Mg of the Earth’s primitive mantle (PM) are from Palme and O’Neill (2014) and Teng et al. (2010), respectively. MgO contents for Martian meteorites are taken from the literature (see main text); for basaltic shergottite NWA 4864, MgO content of the fusion crust of the paired NWA 2975 was taken (see He et al., 2015, and discussion therein). 2SD error bar of ±0.07h was added to (a) and is identical for all panels.

The unusually plagioclase-rich and incompatible element-depleted shergottite NWA 5990 has the highest δ 26 Mg = −0.106 ± 0.058h despite high modal abundance of olivine. There is an overall trend towards heavier Mg isotope compositions with decreasing MgO content (Fig. 2a), i.e. increasing chemical differentiation. Heavier Mg isotope compositions are associated with higher modal abundance of pyroxene in nakhlites (Fig. 2b) whereas they are linked to higher modal plagioclase in some shergottites and NWA 7034 (Fig. 2c). This is particularly apparent for the Los Angeles basaltic shergottite with its distinct chemistry and petrogenesis (Rubin et al., 2000). The entire Martian suite displays a general correlation between Mg isotopes and incompatible trace element parameters (Fig. 3), indicating the role of crystal fraction-

Fig. 3. The δ 26 Mg values versus Cr contents (a), Rb contents (b) and chondritenormalized (Anders and Grevesse, 1989) LaN /SmN values (see Magna et al., 2015a, for references).

ation of olivine and/or clinopyroxene, and perhaps incorporation of chemically more evolved surface materials, carrying higher abundance of incompatible elements such as Rb (Fig. 3b). High δ 26 Mg in NWA 5990 is associated with the highest S abundance whereas the isotopically similar NWA 7034 has low bulk sulphur content, precluding direct links to sulphur-rich Martian surface. 4. Discussion A general problem with analyses of meteorite finds, which represent the majority of the samples, is a potential contamination during their residence on the Earth’s surface. However, significant terrestrial contamination can be excluded for the suite due to high MgO contents of Martian meteorites compared with typical terrestrial upper crustal materials. In addition, there is a lack of correlation between Mg isotopes and indices of residence in desert areas, such as Ba/La (not shown), although Li isotopes have provided evidence for contamination in a sub-set of Martian meteorites during residence on Earth’s surface (e.g., DaG 476, Dhofar 019, perhaps NWA 1068; Magna et al., 2015a). The lack of resolved

JID:EPSL

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.5 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

5

terrestrial contamination is further supported by δ 26 Mg values that are internally consistent regardless whether the samples are falls or finds (e.g., Shergotty and Zagami versus other basaltic shergottites, Tissint versus other olivine-phyric shergotittes, Chassigny versus NWA 2737, Nakhla versus other nakhlites). Even the terrestrially altered sample DaG 476 (Crozaz and Wadhwa, 2001) has a δ 26 Mg value consistent with those of the other depleted shergottites (Table 1), supporting the interpretation that all samples still record their original Mg isotope signature. 4.1. Assessment of the Mg isotope composition of the Martian mantle The scant literature data for Mg isotope compositions of Martian meteorites (Bizzarro et al., 2011; Chakrabarti and Jacobsen, 2010; Wiechert and Halliday, 2007) does not allow for a detailed and robust assessment of the δ 26 Mg value of the Martian mantle. Two earlier analyses of the incompatible element-enriched shergottites Zagami (Wiechert and Halliday, 2007) and NWA 856 (Bizzarro et al., 2011) are in the range of terrestrial peridotites, and mid-ocean ridge and oceanic island basalts (e.g., Bourdon et al., 2010; Teng et al., 2010). This observation would, at first glance, provide evidence for similarity and homogeneity of Mg isotopes in the mantles of Earth and Mars. However, considering that only two enriched basaltic shergottites were used for the estimate, this interpretation may not be robust. A number of studies indicate that basaltic shergottites are not an ideal reservoir to constrain the composition of the Martian mantle (e.g., Nyquist et al., 2009). Their generally low MgO contents (<12.3 wt.%; Table 1), compared with those of olivine-, olivine–orthopyroxene-phyric and lherzolitic shergottites, suggest that petrogenesis of basaltic shergottites involved a number of processes, which included assimilation of chemically evolved crustal lithologies as well as extensive fractional crystallization (Rubin et al., 2000). On the contrary, chassignites (olivine cumulates) may anchor the δ 26 Mg systematics of the Martian mantle more faithfully because Mg isotope fractionation between olivine and melt is small (e.g., Young et al., 2009) and thus the early crystallized olivine that constitutes a significant proportion of these samples will dominate and buffer the Mg budget of the Martian mantle. The importance of olivine during the entire magmatic evolution of Mars is further supported by experimental and modeling constraints (e.g., Elkins-Tanton et al., 2005). Moreover, Mg isotope compositions of olivine seem to be constant over a range of mantle equilibration temperatures (e.g., Handler et al., 2009; Liu et al., 2011; Yang et al., 2009) which underscores that olivine is the key mineral phase for estimating the Mg isotope composition of the accessible part of the Martian mantle. In addition, lherzolitic, olivinephyric and olivine–orthopyroxene-phyric shergottites may closely approach the Mg isotope composition of the Martian mantle given the large olivine proportion and the observation that some of these samples may represent near-primary melts of the Martian mantle (e.g., Y-980459; Usui et al., 2008). By excluding specimens with either Martian crust-related history (NWA 1068; Filiberto et al., 2010), imprint of terrestrial contamination (NWA 4925; Magna et al., 2015a) or other yet unaccounted disturbance (NWA 5990), the following meteorites are used to derive the δ 26 Mg value for the Martian mantle: ALH 77005, Chassigny, DaG 476, EETA 79001A, LAR 06319, NWA 1950, NWA 2737, NWA 6162, RBT 04262, SaU 005–051–094 clan, Tissint, Y-000097, and Y-980459. The cumulative δ 26 Mg of these samples is −0.271 ± 0.040h (2SD, n = 15) and is adopted here to reflect δ 26 Mg value of the Marian mantle; the corresponding δ 25 Mg is −0.133 ± 0.028h (2SD). We note that if chassignites, lherzolitic shergottites and Y-980459 only are considered, an identical δ 26 Mg = −0.271 ± 0.026h (2SD, n = 6) is obtained, with the corresponding δ 25 Mg = −0.132 ± 0.027h

Fig. 4. The range of δ 26 Mg values for various Solar System and Earth reservoirs (after Teng, 2017, and references therein). A general trend toward low δ 26 Mg is observed for reservoirs associated with water. The range for Mars is from this study. AOB – altered oceanic basalts, CC – continental crust.

(2SD). Thus, this newly derived δ 26 Mg value is robust and insensitive to small-degree chemical and mineralogical variability among the samples. This value is similar to that reported for the mantles of the Earth (δ 26 Mg = −0.25 ± 0.04h; Bourdon et al., 2010; Handler et al., 2009; Teng et al., 2010) and Moon (δ 26 Mg = −0.26 ± 0.16h; Sedaghatpour et al., 2013), as well as for chondrites (δ 26 Mg = −0.28 ± 0.06h; e.g., Bourdon et al., 2010; Schiller et al., 2010; Teng et al., 2010) and achondrites (δ 26 Mg = −0.25 ± 0.08h; Handler et al., 2009; Sedaghatpour and Teng, 2016). Notably, the restricted δ 26 Mg range among Martian samples relative to Moon is somewhat surprising when the entire lunar range (−0.61 to −0.14h) is considered, which was advocated to result from the distinctive mineralogy of various lunar reservoirs (Sedaghatpour et al., 2013). Although past wet periods in Martian history compared to a rather dry history of Moon could have imparted sizeable effects on δ 26 Mg, the lack of distinctly light or heavy Mg isotope signatures observed for materials altered at low temperatures (see a recent review of Teng, 2017) indicates that ephemeral aqueous events may have had a limited effect on the planetary-scale Mg cycle and that only intense shallow hydrous alteration may induce significant Mg isotope fractionation. 4.2. Magnesium isotope signature of the Martian crust The surface mineralogy plays an important role in deconvolving the Mg isotope signature of sedimentary materials. Earth’s upper continental crust has an average δ 26 Mg = −0.22 ± 0.10h (Li et al., 2010) but the Mg isotope variability of crustal magmatic and sedimentary materials shows a significant δ 26 Mg spread of >7h (−5.6 to +2.0h) (Fig. 4; Teng, 2017). Carbonates show predominantly light Mg isotope signatures whereas silicate sediments/soils have prevalently heavy Mg isotope systematics although weathering at low-temperature to hydrothermal conditions may give rise to a large range of δ 26 Mg values (e.g., Li et al., 2010; Tipper et al., 2006).

JID:EPSL

6

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.6 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

Fig. 5. Schematic mixing/assimilation trends between Bulk Silicate Mars (BSM) versus putative Martian crust and low-temperature weathering products. A range of MgO contents at the same δ 26 Mg is plotted to accommodate lithologic diversity of Martian meteorites, illite data is from Wimpenny et al. (2014). Mg-sulfates could represent a minor component provided their δ 26 Mg would be known. The δ 26 Mg = 0.5h has been selected for the Martian crust although its value is currently not constrained.

Martian sediments show a range in mineralogy, dominated by phyllosilicates (Fe/Mg/Al smectites and kaolin group minerals) and Cl- and SO4 -bearing minerals, including epsomite, kieserite, gypsum and jarosite, while carbonates are rare on Mars (Ehlmann and Edwards, 2014; Chou and Seal, 2007; McSween, 2015). It was already revealed by the Viking landers that Mg-sulfates are widespread in sedimentary sequences on the Martian surface where they are a key phase in the cement of soils and dust (Toulmin et al., 1977). The Mg isotope fractionation between solutions and sulfates during evaporation is little constrained but appears to be strongly phase-dependent (Geske et al., 2015; Li et al., 2011b). These experimental studies have shown that δ 26 Mg variations between solutions and crystallized phases may be as large as ∼2h. For example, experimentally grown epsomite was ∼0.6h heavier than the solution (Li et al., 2011b) whereas gypsum can be isotopically lighter relative to pore water (Geske et al., 2015). Furthermore, clays may acquire an isotopically heavy signature (Wimpenny et al., 2014) and evidence for the former presence of phyllosilicates was reported for NWA 7533 (paired with NWA 7034; Humayun et al., 2014). The isotopically ‘heavy’ Mg in NWA 7034 can be compared with the black-glass fraction of Tissint which also tends toward higher δ 26 Mg (Table 1). This fraction has been reported to contain trapped Martian atmospheric C and N, as well as an embedded Martian soil component, as indicated by minor (F, S) and trace element (REE) systematics (Chennaoui Aoudjehane et al., 2012). This is consistent with isotopically fractionated Li signature found for black glass compared to bulk Tissint, which may reflect incorporation of surface materials that have formed at low temperatures (Magna et al., 2015a). A schematic model of mixing of isotopically heavier materials is presented in Fig. 5 where data for illite is used (Wimpenny et al., 2014) but we note here that this model is strongly end-member dependent. Indeed, only ALH 84001 and black glass fraction of Tissint show an imprint from such materials where Mg-sulfates could also be a viable assimilate, but in the absence of experimental data no further constraints can be made. Crustal breccia NWA 7034 appears to have developed its marginally heavier δ 26 Mg via other mechanisms because of its low MgO content relative to most other Martian meteorites, which would require a very low MgO content in its source, extremely high δ 26 Mg of the contaminant and/or loss of isotopically light Mg during weathering. At present, it is unclear whether weathering-associated minerals or percolating fluids/brines are responsible for isotopi-

Fig. 6. The δ 26 Mg versus δ 7 Li (a) and δ 37 Cl (b) in Martian samples. Lithium and Cl isotope data are taken from elsewhere (Magna et al., 2015a; Williams et al., 2016). Grey bars in (a) represent the Mg–Li isotope ranges for Bulk Silicate Mars (BSM). Star in (b) represents BSM value for Mg (this study) and Cl isotope composition (Williams et al., 2016), respectively. Grey arrow indicates possible direction towards a putative Martian crustal reservoir.

cally shifted Mg in surface-related lithologies. In this respect, Librine /Lirock is most likely much higher than Mgbrine /Mgrock , indicating that percolating brines may have a more limited effect on the Mg isotope systematics of bulk rocks compared to Li. The apparent scatter in Fig. 6a suggests that Li and Mg are affected by distinct processes. However, it is notable that isotopically fractionated Mg (and Li) in NWA 7034, coupled with the chemical match of this meteorite with remote sensing and in-situ data for the Martian surface (McSween, 2015), currently provides the best δ 26 Mg (and δ 7 Li) estimate for the Martian crust. It is also apparent that some of the heavy Mg isotope values (including those in nakhlites) are paralleled by isotopically heavy Cl (Fig. 6b) where elevated δ 37 Cl has been assigned to crust–fluid interaction (Williams et al., 2016). Collectively, it is still difficult to provide a more qualitative estimate of the Mg isotope systematics of the chemically evolved silicate portion of Mars and percolating fluids at Martian atmospheric and hydrologic conditions. Only careful experimental studies may constrain the Mg isotope systematics of crustal materials, including specific compounds such as chlorides and sulfates, which can potentially be an important carrier of isotopically fractionated Mg. Also, the knowledge of Mg isotope composition of past water masses would be critical in estimating the surface Mg signature.

JID:EPSL

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.7 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

4.3. ‘Heavy’ Mg isotopes in nakhlites The generally higher δ 26 Mg in nakhlites (>−0.195 ± 0.072h), compared with most shergottites (in particular olivine-phyric, olivine–orthopyroxene-phyric and lherzolitic shergottites), is intriguing as it could imply fractionation of Mg isotopes during a magmatic process (see also Schiller et al., 2017). However, there are three different processes that could generate the observed elevated δ 26 Mg values (Table 1; Fig. 1). (i) Clinopyroxene is isotopically heavier compared to olivine and orthopyroxene in the mantle (Wiechert and Halliday, 2007; Xiao et al., 2013; Young et al., 2009) and a high modal proportion of clinopyroxene in nakhlites makes it a principal candidate for high δ 26 Mg values. We note that Lafayette, as the most equilibrated nakhlite from this study (Treiman, 2005), has the lowest δ 26 Mg value of −0.195 ± 0.072h which would be the closest to the source signature of nakhlites. However, some shergottites (EETA 79001A, NWA 856, Shergotty, Zagami) have high modal contents of pyroxene but their δ 26 Mg values are in the range of the Martian mantle (−0.261 ± 0.069h to −0.193 ± 0.065h) and overlap only marginally with those of nakhlites (−0.195 ± 0.072h to −0.113 ± 0.072h; Fig. 2b). Moreover, some Martian meteorites also have a heavy Mg isotope signature while, at the same time, they have a different mineralogy. This can be illustrated using Mg isotope data for the highly differentiated shergottite Los Angeles (Rubin et al., 2000). Its δ 26 Mg of −0.108 ± 0.065h and the lowest MgO content (3.5 wt.%; Table 1) among all Martian meteorites likely result from extensive magmatic differentiation, reflected by its low clinopyroxene content (Fig. 2b) and the distinctly high proportion of plagioclase (Fig. 2c) compared with other basaltic shergottites. (ii) Debaille et al. (2009) provided evidence for garnet segregation from the mantle source of nakhlites (also Draper et al., 2003, for example). Garnet has been shown to preferentially incorporate light Mg isotopes (e.g., Huang et al., 2013; Li et al., 2011a) which translates into isotopically heavy residual melts. In the absence of fractionation factors for Mg isotopes relevant to Martian conditions and considering both the cumulate nature of nakhlites and a possibility of a direct genetic disconnect between some nakhlites (Jambon et al., 2016), any straightforward crystal fractionation model would be poorly constrained by relevant quantitative experimental data for Mg isotope fractionation. (iii) Assimilation of surface materials has been advocated for shallow-seated nakhlites such as MIL 03346 (Day et al., 2006; Magna et al., 2015a) and Mg-sulfates have been reported to occur in Nakhla (Bridges and Grady, 2000). Interestingly, Nakhla has the highest δ 26 Mg value of all measured nakhlites (δ 26 Mg = −0.113h), which could in part be due to these late-stage alteration features on Mars. The current data set provides indirect evidence for isotopically heavy Mg in the Martian crust compared to its mantle (Sections 4.1 and 4.2), although the exact magnitude and the cause of the δ 26 Mg difference remain unconstrained at present. Notably, Mg isotope systematics do not support the idea of large-scale assimilation of non-silicate parts of the Martian crust (i.e., Mg-sulfates and -chlorides, etc.). Such contamination would also increase MgO of the nakhlites, which is not observed (Fig. 5). On the other hand, Li isotopes indicate crustal contamination in a subset of samples (Magna et al., 2015a). More quantitative constraints can thus only be made once the variability and the extent of Mg isotope fractionation in Martian surface lithologies and sedimentary deposits are known. Currently, clinopyroxene accumulation appears to best explain the observed Mg isotope systematics of nakhlites and indicates that surface contamination had only a marginal effect on Mg isotope systematics.

7

5. Conclusions The new Mg isotope data for a suite of Martian meteorites display a limited variation in δ 26 Mg values (−0.318 to −0.106h), irrespective of a large range in lithology and distinct petrogenesis. None of the primitive Martian samples show resolvable deviation from the estimated bulk Martian mantle δ 26 Mg value at −0.271 ± 0.040h (2SD). This is consistent with the observation that all rocky planetesimals have similar Mg isotope composition as bulk Earth and Moon at δ 26 Mg of −0.28h. This Mg isotope uniformity also implies a well-mixed reservoir of the major Mg-bearing phases in the protoplanetary disk. This homogeneous distribution of Mg isotopes beyond the currently accepted analytical resolution contrasts with the observation that Si shows mass-dependent isotope fractionation among different Solar System bodies, which may reflect mass-dependent fractionation during nebular condensation of forsterite or planetesimal collisions (Dauphas et al., 2015; Pringle et al., 2014). At present, it is unclear whether or not early silicate differentiation of Mars imparted measurable effects on Mg isotopes such that all δ 26 Mg variations observed could unequivocally be attributed to magmatic differentiation during the genesis of SNC meteorites. Since Mg isotopes fractionate readily during low-temperature processes it could be expected that Mg-bearing minerals that formed during fluid–rock interaction and weathering on the surface of Mars have distinctive Mg isotope signatures. This appears to have possibly been detected in crustal breccia NWA 7034 and black glass of the Tissint shergottite, although the exact origin of isotopically heavy Mg at the Martian surface remains unclear. Alternatively, the slightly heavier Mg isotope composition of some evolved materials may also be attributed to crystal fractionation during solidification of the melts. The homogeneous Mg isotope composition of the Martian mantle would, therefore, record a lack of large-scale recycling of isotopically fractionated surface materials into the source of silicate melts that formed the precursors of the SNC meteorites. Careful experimental work could provide further constraints on these intriguing possibilities, with far-reaching implications for the dynamic evolution of Mars. Acknowledgements We are grateful to NASA, Tim McCoy (Smithsonian Institution), Addi Bischoff (Universität Münster), Franz Brandstätter (NHM Vienna), Beda Hofmann (Universität Bern), Hideyasu Kojima (NIPR), Jutta Zipfel (Senckenberg Naturmuseum Frankfurt) and Norbert Clausen for sample allocations. We express our special gratitude to Carl Agee and Hasnaa Chennaoui Aoudjehane for allocation of some samples for the previous work. We thank Martin Bizzarro for additional information on Martian Mg data, Hap McSween for information on Martian petrogenesis and Andreas Stracke for discussion. This work has been funded by the Helmholtz Association through the research alliance HA 203 “Planetary Evolution and Life”. TM acknowledges a partial support by the Czech Science Foundation project 13-22351S. FZT is supported by National Science Foundation (EAR-1340160). KM acknowledges partial support through the NCCR PlanetS. We are grateful to Martin Schiller for in-depth review and editorial handling by Frédéric Moynier. References Agee, C.B., Wilson, N.V., McCubbin, F.M., Ziegler, K., Polyak, V.J., Sharp, Z.D., Asmerom, Y., Nunn, M.H., Shaheen, R., Thiemens, M.H., Steele, A., Fogel, M.L., Bowden, R., Glamoclija, M., Zhang, Z., Elardo, S.M., 2013. Unique meteorite from Early Amazonian Mars: water-rich basaltic breccia Northwest Africa 7034. Science 339, 780–785. Anders, E., Grevesse, N., 1989. Abundances of the elements: meteoritic and solar. Geochim. Cosmochim. Acta 53, 197–214. Bizzarro, M., Paton, C., Larsen, K., Schiller, M., Trinquier, A., Ulfbeck, D., 2011. Highprecision Mg-isotope measurements of terrestrial and extraterrestrial material

JID:EPSL

8

AID:14554 /SCO

[m5G; v1.221; Prn:26/07/2017; 8:07] P.8 (1-8)

T. Magna et al. / Earth and Planetary Science Letters ••• (••••) •••–•••

by HR-MC-ICPMS—implications for the relative and absolute Mg-isotope composition of the bulk silicate Earth. J. Anal. At. Spectrom. 26, 565–577. Bourdon, B., Tipper, E.T., Fitoussi, C., Stracke, A., 2010. Chondritic Mg isotope composition of the Earth. Geochim. Cosmochim. Acta 74, 5069–5083. Breuer, D., Spohn, T., 2003. Early plate tectonics versus single-plate tectonics on Mars: evidence from magnetic field history and crust evolution. J. Geophys. Res., Planets 108, 5072. http://dx.doi.org/10.1029/2002JE001999. Bridges, J.C., Grady, M.M., 2000. Evaporite mineral assemblages in the nakhlite (Martian) meteorites. Earth Planet. Sci. Lett. 176, 267–279. Chakrabarti, R., Jacobsen, S.B., 2010. The isotopic composition of magnesium in the inner Solar System. Earth Planet. Sci. Lett. 293, 349–358. Chennaoui Aoudjehane, H., Avice, G., Barrat, J.A., Boudouma, O., Chen, G., Duke, M.J.M., Franchi, I.A., Gattacceca, J., Grady, M.M., Greenwood, R.C., Herd, C.D.K., Hewins, R., Jambon, A., Marty, B., Rochette, P., Smith, C.L., Sautter, V., Verchovsky, A., Weber, P., Zanda, B., 2012. Tissint martian meteorite: a fresh look at the interior, surface, and atmosphere of Mars. Science 338, 785–788. Chou, I.-M., Seal II, R.R., 2007. Magnesium and calcium sulfate stabilities and the water budget of Mars. J. Geophys. Res. 112, E11004. http://dx.doi.org/10.1029/ 2007JE002898. Crozaz, G., Wadhwa, M., 2001. The terrestrial alteration of Saharan shergottites Dar al Gani 476 and 489: a case study of weathering in a hot desert environment. Geochim. Cosmochim. Acta 65, 971–978. Dauphas, N., Poitrasson, F., Burkhardt, C., Kobayashi, H., Kurosawa, K., 2015. Planetary and meteoritic Mg/Si and δ 30 Si variations inherited from solar nebula chemistry. Earth Planet. Sci. Lett. 427, 236–248. Day, J.M.D., Taylor, L.A., Floss, C., McSween, H.Y., 2006. Petrology and chemistry of MIL 03346 and its significance in understanding the petrogenesis of nakhlites on Mars. Meteorit. Planet. Sci. 41, 581–606. Debaille, V., Brandon, A.D., O´Neill, C., Yin, Q.-Z., Jacobsen, B., 2009. Early martian mantle overturn inferred from isotopic composition of nakhlite meteorites. Nat. Geosci. 2, 548–552. Draper, D.S., Xirouchakis, D., Agee, C.B., 2003. Trace element partitioning between garnet and chondritic melt from 5 to 9 GPa: implications for the onset of the majorite transition in the martian mantle. Phys. Earth Planet. Inter. 139, 149–169. Ehlmann, B.L., Edwards, C.S., 2014. Mineralogy of the martian surface. Annu. Rev. Earth Planet. Sci. 42, 291–315. Elkins-Tanton, L.T., Hess, P.C., Parmentier, E.M., 2005. Possible formation of ancient crust on Mars through magma ocean processes. J. Geophys. Res. 110, E12S01. http://dx.doi.org/10.1029/2005JE002480. Filiberto, J., Musselwhite, D.S., Gross, J., Burgess, K., Le, L., Treiman, A.H., 2010. Experimental petrology, crystallization history, and parental magma characteristics of olivine-phyric shergottite NWA 1068: implications for the petrogenesis of “enriched” olivine-phyric shergottites. Meteorit. Planet. Sci. 45, 1258–1270. Galy, A., Yoffe, O., Janney, P.E., Williams, R.W., Cloquet, C., Alard, O., Halicz, L., Wadhwa, M., Hutcheon, I.D., Ramon, E., Carignan, J., 2003. Magnesium isotope heterogeneity of the isotopic standard SRM980 and new reference materials for magnesium-isotope-ratio measurements. J. Anal. At. Spectrom. 18, 1352–1356. Geske, A., Lokier, S., Dietzel, M., Richter, D.K., Buhl, D., Immenhauser, A., 2015. Magnesium isotope composition of sabkha porewater and related (sub-)recent stoichiometric dolomites, Abu Dhabi (UAE). Chem. Geol. 393–394, 112–124. Handler, M.R., Baker, J.A., Schiller, M., Bennett, V.C., Yaxley, G.A., 2009. Magnesium stable isotope composition of Earth’s upper mantle. Earth Planet. Sci. Lett. 282, 306–313. He, Q., Xiao, L., Balta, J.B., Baziotis, I.P., Hsu, W., Guan, Y., 2015. Petrography and geochemistry of the enriched basaltic shergottite Northwest Africa 2975. Meteorit. Planet. Sci. 50, 2024–2044. Huang, F., Chen, L., Wu, Z., Wang, W., 2013. First-principles calculations of equilibrium Mg isotope fractionations between garnet, clinopyroxene, orthopyroxene, and olivine: implications for Mg isotope thermometry. Earth Planet. Sci. Lett. 367, 61–70. Huang, J.-X., Xiang, Y.X., An, Y., Griffin, W.L., Gréau, Y., Xie, L., Pearson, N.J., Yu, H., O’Reilly, S.Y., 2016. Magnesium and oxygen isotopes in Roberts Victor eclogites. Chem. Geol. 438, 73–83. Humayun, M., Hewins, R.H., Zanda, B., 2014. Evidence for phyllosilicate precursors in an aphanitic spherule from NWA 7533. Annu. Meteorit. Soc. Meeting 77, #5413. Jambon, A., Sautter, V., Barrat, J.-A., Gattacceca, J., Rochette, P., Boudouma, O., Badia, D., Devouard, B., 2016. Northwest Africa 5790: revisiting nakhlite petrogenesis. Geochim. Cosmochim. Acta 190, 191–212. Ke, S., Teng, F.-Z., Li, S.-G., Gao, T., Liu, S.-A., He, Y., Mo, X., 2016. Mg, Sr, and O isotope geochemistry of syenites from northwest Xinjiang, China: tracing carbonate recycling during Tethyan oceanic subduction. Chem. Geol. 437, 109–119. Li, W.-Y., Teng, F.-Z., Ke, S., Rudnick, R.L., Gao, S., Wu, F.-Y., Chappell, B.W., 2010. Heterogeneous magnesium isotopic composition of the upper continental crust. Geochim. Cosmochim. Acta 74, 6867–6884. Li, W.-Y., Teng, F.-Z., Xiao, Y., Huang, J., 2011a. High-temperature inter-mineral magnesium isotope fractionation in eclogite from the Dabie orogen, China. Earth Planet. Sci. Lett. 304, 224–230. Li, W., Beard, B.L., Johnson, C.M., 2011b. Exchange and fractionation of Mg isotopes between epsomite and saturated MgSO4 solution. Geochim. Cosmochim. Acta 74, 1814–1828.

Liu, S.-A., Teng, F.-Z., Yang, W., Wu, F.-Y., 2011. High-temperature inter-mineral magnesium isotope fractionation in mantle xenoliths from the North China craton. Earth Planet. Sci. Lett. 308, 131–140. Magna, T., Day, J.M.D., Mezger, K., Fehr, M.A., Dohmen, R., Chennaoui Aoudjehane, H., Agee, C., 2015a. Lithium isotope constraints on crust–mantle interactions and surface processes on Mars. Geochim. Cosmochim. Acta 162, 46–65. Magna, T., Gussone, N., Mezger, K., 2015b. The calcium isotope systematics of Mars. Earth Planet. Sci. Lett. 430, 86–94. McSween, H.Y., 2015. Petrology on Mars. Am. Mineral. 100, 2380–2395. McSween, H.Y., McLennan, S.M., 2014. Mars. In: Davis, A.M. (Ed.), Treatise on Geochemistry, 2nd edn. Elsevier Ltd., Oxford, UK, pp. 1–27. Nyquist, L.E., Bogard, D.D., Shih, C.-Y., Park, J., Reese, Y.D., Irving, A.J., 2009. Concordant Rb–Sr, Sm–Nd, and Ar–Ar ages for Northwest Africa 1460: a 346 Ma old basaltic shergottite related to “lherzolitic” shergottites. Geochim. Cosmochim. Acta 73, 4288–4309. Palme, H., O’Neill, H.S.C., 2014. Cosmochemical estimates of mantle composition. In: Carlson, R.W. (Ed.), Treatise on Geochemistry, 2nd edition. Elsevier Ltd., Oxford, UK, pp. 1–39. Pringle, E.A., Moynier, F., Savage, P.S., Badro, J., Barrat, J.-A., 2014. Silicon isotopes in angrites and volatile loss in planetesimals. Proc. Natl. Acad. Sci. 111, 17029–17032. Rubin, A.E., Warren, P.H., Greenwood, J.P., Verish, R.S., Leshin, L.A., Hervig, R.L., Clayton, R.N., Mayeda, T.K., 2000. Los Angeles: the most differentiated basaltic martian meteorite. Geology 28, 1011–1014. Sedaghatpour, F., Teng, F.-Z., 2016. Magnesium isotopic composition of achondrites. Geochim. Cosmochim. Acta 174, 167–179. Sedaghatpour, F., Teng, F.-Z., Liu, Y., Sears, D.W.G., Taylor, L.A., 2013. Magnesium isotopic composition of the Moon. Geochim. Cosmochim. Acta 120, 1–19. Schiller, M., Dallas, J.A., Creech, J., Bizzarro, M., Baker, J.A., 2017. Tracking the formation of magma oceans in the Solar System using stable magnesium isotopes. Geochem. Persp. Lett. 3, 22–31. Schiller, M., Handler, M., Baker, J.A., 2010. High-precision Mg isotopic systematics of bulk chondrites. Earth Planet. Sci. Lett. 297, 165–173. Sossi, P.A., Nebel, O., Anand, M., Poitrasson, F., 2016. On the iron isotope composition of Mars and volatile depletion in the terrestrial planets. Earth Planet. Sci. Lett. 449, 360–371. Teng, F.-Z., 2017. Magnesium isotope geochemistry. Rev. Mineral. Geochem. 82, 219–287. Teng, F.-Z., Li, W.-Y., Ke, S., Marty, B., Dauphas, N., Huang, S., Wu, F.-Y., Pourmand, A., 2010. Magnesium isotopic composition of the Earth and chondrites. Geochim. Cosmochim. Acta 74, 4150–4166. Teng, F.-Z., Li, W.-Y., Ke, S., Yang, W., Liu, S.-A., Sedaghatpour, F., Wang, S.-J., Huang, K.-J., Hu, Y., Ling, M.-X., Xiao, Y., Liu, X.-M., Li, X.-W., Gu, H.-O., Sio, C.K., Wallace, D.A., Su, B.-X., Zhao, L., Chamberlin, J., Harrington, M., Brewer, A., 2015. Magnesium isotopic compositions of international geological reference materials. Geostand. Geoanal. Res. 39, 329–339. Tipper, E.T., Galy, A., Bickle, M.J., 2006. Riverine evidence for a fractionated reservoir of Ca and Mg on the continents: implications for the oceanic Ca cycle. Earth Planet. Sci. Lett. 247, 267–279. Toulmin III, P., Baird, A.K., Clark, B.C., Keil, K., Rose, H.J., Christian, R.P., Evans, P.H., Kelliher, W.C., 1977. Geochemical and mineralogical interpretation of the Viking inorganic chemical results. J. Geophys. Res. 82, 4625–4634. Treiman, A.H., 2005. The nakhlite meteorites: augite-rich igneous rocks from Mars. Chem. Erde 65, 203–270. Usui, T., McSween, H.Y., Floss, C., 2008. Petrogenesis of olivine–phyric shergottite Yamato 980459, revisited. Geochim. Cosmochim. Acta 72, 1711–1730. Valdes, M.C., Moreira, M., Foriel, J., Moynier, F., 2014. The nature of Earth’s building blocks as revealed by calcium isotopes. Earth Planet. Sci. Lett. 394, 135–145. Wiechert, U., Halliday, A.N., 2007. Non-chondritic magnesium and the origins of the inner terrestrial planets. Earth Planet. Sci. Lett. 256, 360–371. Williams, J.T., Shearer, C.K., Sharp, Z.D., Burger, P.V., McCubbin, F.M., Santos, A.R., Agee, C.B., McKeegan, K.D., 2016. The chlorine isotopic composition of Martian meteorites 1: chlorine isotope composition of Martian mantle and crustal reservoirs and their interactions. Meteorit. Planet. Sci. 51, 2092–2110. Wimpenny, J., Colla, C.A., Yin, Q.-Z., Rustad, J.R., Casey, W.H., 2014. Investigating the behaviour of Mg isotopes during the formation of clay minerals. Geochim. Cosmochim. Acta 128, 178–194. Xiao, Y., Teng, F.-Z., Zhang, H.-F., Yang, W., 2013. Large magnesium isotope fractionation in peridotite xenoliths from eastern North China craton: product of melt–rock interaction. Geochim. Cosmochim. Acta 115, 241–261. Yang, W., Teng, F.-Z., Li, W.-Y., Liu, S.-A., Ke, S., Liu, Y.-S., Zhang, H.-F., Gao, S., 2016. Magnesium isotopic composition of the deep continental crust. Am. Mineral. 101, 243–252. Yang, W., Teng, F.-Z., Zhang, H.-F., 2009. Chondritic magnesium isotopic composition of the terrestrial mantle: a case study of peridotite xenoliths from the North China craton. Earth Planet. Sci. Lett. 288, 475–482. Young, E.D., Tonui, E., Manning, C.E., Schauble, E., Macris, C.A., 2009. Spinel–olivine magnesium isotope thermometry in the mantle and implications for the Mg isotopic composition of Earth. Earth Planet. Sci. Lett. 288, 524–533.