Earth and Planetary Science Letters 389 (2014) 167–175

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

www.elsevier.com/locate/epsl

Vanadium isotopic difference between the silicate Earth andmeteorites

Sune G. Nielsen a,b,∗, Julie Prytulak a,c, Bernard J. Wood a, Alex N. Halliday a

a Department of Earth Sciences, University of Oxford, South Parks Road, OX1 3AN, Oxford, UKb Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USAc Department of Earth Science and Engineering, Imperial College London, SW7 2AZ, UK

a r t i c l e i n f o a b s t r a c t

Article history:Received 30 April 2013Received in revised form 14 December 2013Accepted 20 December 2013Available online 14 January 2014Editor: T. Elliott

Keywords:solar system irradiationchondritic earthvanadium isotopesachondriteschondrites

It has been argued that the stable isotopic composition of the element vanadium (V) provides a potentialindicator of the effects high-energy irradiation early in Solar System development. Such irradiationwould produce enrichment in the minor isotope, 50V compared with the 400 times more abundant 51V(Gounelle et al., 2001; Lee et al., 1998). Here we show that the vanadium isotopic composition of thesilicate Earth is enriched in 51V by ∼0.8� compared with carbonaceous and ordinary chondrites as wellas achondrites from Mars and the asteroid 4 Vesta. Although V is depleted by core formation, experimentsreveal no isotopic fractionation between metal and silicate that could account for the observed differencein V isotope composition between terrestrial and extraterrestrial materials. Nucleosynthetic provenanceof the terrestrial vanadium isotope offset is inconsistent with anomalies of other nucleosyntheticallyproduced isotopes in bulk meteorites, which are more variable than vanadium (Burkhardt et al., 2011;Carlson et al., 2007; Trinquier et al., 2009). Furthermore, V isotopes are unlikely to have been affectedby volatilization, parent body alteration or impact erosion of Earth’s surface. Therefore, the cause of theisotopic difference is unclear. One possibility is that Earth’s isotopically heavier V reflects a deficit inmaterial irradiated during the initial stages of Solar System formation. Whatever the cause, the terrestrialdeficit in 50V implies that bulk Earth cannot be entirely reconstructed by mixtures of different meteorites.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

It is believed that the earliest phases of the Sun’s evolutionwere characterized by frequent violent bursts of high-energy par-ticles (Shu et al., 1997). However, the consequences for the syn-chronous formation and distribution of the first solids in the So-lar System are the topic of much debate. In a series of papersShu and co-workers developed a physical model in which cal-cium aluminium rich inclusions (CAIs) and possibly chondruleswere formed close to the Sun in a high-energy irradiation envi-ronment and subsequently transported to the Asteroid Belt via theso-called X-wind (Lee et al., 1998; Shu et al., 1996, 1997, 2001).The X-wind is a specific type of magnetocentrifugal outflow of ma-terial, which is a common feature of young stellar objects (Coffeyet al., 2007). It was also proposed that a host of short-lived ra-dioactive isotopes, such as 10Be, 26Al and 41Ca, would have formedin this environment, which was suggested could explain the ob-served abundances of these isotopes in CAIs (Gounelle et al., 2001;

* Corresponding author at: Department of Geology and Geophysics, Woods HoleOceanographic Institution, Woods Hole, MA 02543, USA.

E-mail address: snielsen@whoi.edu (S.G. Nielsen).

0012-821X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.epsl.2013.12.030

Lee et al., 1998). However, several lines of evidence indicatethat the specific physical mechanisms invoked for formation andtransport of CAIs, chondrules and short-lived radioactive nuclides(SLRs) are not realistic in an X-wind scenario (Desch et al., 2010;Wood, 2004).

Still, there is compelling evidence to indicate that irradia-tion was an integral part of the early stages of Solar Systemevolution both based on astronomical and meteoritic observa-tions (Burnett et al., 1965; Coffey et al., 2007; Feigelson, 2010;Fowler et al., 1962; Jacobsen et al., 2011; McKeegan et al., 2000).Therefore, more information is required before we can obtain abetter understanding of the processes that shaped the compositionof our Solar System.

The aforementioned calculations of the isotopic consequencesof the X-wind model also considered the stable isotope system ofvanadium (Gounelle et al., 2001; Lee et al., 1998). Even though theX-wind model is difficult to reconcile with our knowledge of CAIs,chondrules and some SLRs, the effects of Solar System irradiationis still relevant for predicting potential isotopic anomalies in irradi-ated material. It was shown that significant enrichments of severalpermil in 50V would be expected for material processed in anenvironment with high-energy irradiation (Gounelle et al., 2001;Lee et al., 1998). Therefore, high-precision measurements of V

168 S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175

isotopes in extraterrestrial material may reveal new informationabout the irradiation history of the Solar System.

Here we present the first measurements of the vanadium iso-topic composition of a number of chondrites and achondrites fromMars and the HED parent body. Our results reveal a striking vana-dium isotopic offset between the Earth and extraterrestrial mate-rial, which is most consistent with Earth accreting from materialthat was less irradiated than Mars and the Asteroid Belt. Thisfinding supports the hypothesis that Earth is not made entirelyfrom material with chondritic relative abundances of refractory el-ements.

2. Vanadium isotope background

The element vanadium (V) has two isotopes with masses 50and 51 and abundances of 0.24% and 99.76%, respectively. The iso-tope composition of vanadium has long attracted interest fromcosmochemists because it was realized that high-energy irradia-tion has significant potential to produce 50-V mainly from targetnuclei of Ti and Cr (Burnett et al., 1965). It was therefore be-lieved that V isotope compositions of meteorites and lunar samplescould elucidate the irradiation history of the early solar systemand potentially show if irradiated material was heterogeneouslydistributed within the solar system. However, despite significantanalytical endeavors no researchers were able to identify any Visotopic variation outside an analytical uncertainty of about ±1%(Balsiger et al., 1969, 1976; Pelly et al., 1970).

Recently the first method that produces highly precise and ac-curate V isotope compositions to an external two standard devi-ations precision of ±0.015% was developed (Nielsen et al., 2011;Prytulak et al., 2011). This advance in analytical precision is almosttwo orders of magnitude superior to earlier efforts and allows thefirst high-precision investigation of vanadium isotopic variation inmeteorites and on Earth. Vanadium isotope compositions are re-ported as

δ51V = 1000 × [(51V/50Vsample − 51V/50VAA)/51V/50VAA

]

where AA is a V solution standard purchased from Alfa Aesar(Lot #91-092043G) which is defined as δ51V = 0. Thus negativevalues denote enrichments in 50V.

3. Methods

3.1. Vanadium isotopic measurements

Vanadium isotope data were obtained using previously de-scribed chemical separation and mass spectrometry methods(Nielsen et al., 2011; Prytulak et al., 2011). About 3–5 μg of V wasconsumed in each individual isotopic measurement and thereforearound 50–100 mg of sample was weighed out for each samplesplit in order to have enough V for at least one measurement.Samples were digested in Teflon vials in mixtures of HF and HNO3either on a hotplate (achondrites) or in an Anton Parr microwavedigestion system (chondrites). These procedures ensured completedigestion of all samples and no residues were encountered for anyof the samples investigated, except for the ureilite where minoramounts of graphite were still present after microwave digestion.These residues were removed prior to V separation via centrifuga-tion as it was assumed that graphite contains negligible amountsof V. Briefly, chemical separation is performed with liquid anionexchange chromatography where each sample is subjected to atleast five column passes. The procedure is especially designed toremove the elements Cr and Ti quantitatively, because 50Cr and50Ti interfere directly on 50V during mass spectrometry and aredifficult to correct for when either 53Cr/51V or 49Ti/51V > 0.00001

Fig. 1. Standard addition tests for Allende and USGS Icelandic basalt BIR-1a in whichsmall known amounts of isotopically fractionated standard VISSOX was added tosplits of Allende and BIR-1a (Prytulak et al., 2011). Regression lines with pureVISSOX and VISSOX-sample mixtures yield intercepts that are identical to unspikedsamples. This coherence shows that no detectable matrix effects are present whenmeasuring chondrites or basalts, which encompass all sample compositions investi-gated in this study. Error bars are two standard deviations on repeat measurementsof the same solution. Symbols without error bars have uncertainties smaller thanthe symbol size.

(Nielsen et al., 2011). All samples run during the course of thisstudy exhibited 53Cr/51V or 49Ti/51V < 0.00001, most with valuesbeing below 0.000003. The procedure returns quantitative yieldsfor V and thus eliminates any potential V isotope fractionationduring chemical separation. Chemical yield of V during columnchemistry was monitored throughout this study by comparing Vconcentrations (measured on minor splits by ICP-MS) with theamount of V recovered from the column chemistry procedure. Allsamples were found to have yields of 100% ± 15%. Isotope ratioswere measured on a Nu Instruments MC-ICPMS relative to an AlfaAesar specpure standard (Nielsen et al., 2011).

Several tests were performed in order to assess precision andaccuracy of V isotopic measurements on both silicate rocks andchondrites. Specifically, we conducted standard addition tests onAllende and USGS Iceland basalt BIR-1a in which known amountsof an isotopically fractionated standard (VISSOX: Vanadium IsotopeStandard Solution OXford) was added in different proportions tosplits of the samples. These standard-sample mixtures were thenprocessed as regular samples and isotope compositions were mea-sured. Fig. 1 shows the mixing lines produced by the standardaddition data and how the intercepts of these mixing lines com-pare with the unspiked samples. It is evident that unspiked sam-ples return values that are identical within error of the mixing lineintercepts, which shows that, within the uncertainties of our mea-surements, there are no systematic matrix effects or inaccuraciesthat affected the isotope compositions of the meteorites (or terres-trial samples published elsewhere by Prytulak et al., 2013).

Precision was determined via multiple measurements of thesame sample where separate splits of the powdered sample weredigested and processed through chemistry and then measured in-dividually on the mass spectrometer. In order to detect inconsis-tencies, all samples in this study were analyzed at least in dupli-cate and the reproducibility is similar to the long-term precision ofUSGS reference rock samples and Allende of about 0.1–0.2� (Ta-ble 1) (Prytulak et al., 2011, 2013).

S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175 169

Table 1Vanadium isotope composition of standards analyzed during each analytical session.

Session date Samples analyzed Mean δ51V BDH 2sd n

Nov. 4–5 2008 Allende −1.20 0.07 32Jan. 27–31 2009 ALH83100.202; ALH83108.40; EET96042.18; Plainview −1.21 0.09 55Feb. 20–22 2009 EET96042.18; Plainview −1.22 0.10 31Mar. 6–10 2009 EET92002.15; Allende; ALH83100.202; ALH83108.40 −1.19 0.10 76Apr. 23–25 2009 Johnstown; Pasamonte; Experiments −1.20 0.09 55Nov. 2–4 2009 Pasamonte; Kapoeta; Orgueil; Plainview −1.20 0.16 39Feb. 9–12 2010 Experiments; Nakhla −1.17 0.10 87Feb. 20–21 2010a Kapoeta; Johnstown; Orgueil; Nakhla −1.24 0.06 35

a Measurements performed at Imperial College London.

3.2. Metal–silicate partitioning experiments

All starting materials for high-pressure experiments were basedon an approximately 1:1 mix of Fe metal and silicate. The com-position of the latter was chosen to be close to the 1.5 GPaeutectic composition in the system anorthite–diopside–forsterite(An50Di28Fo22) (Presnall et al., 1978). Approximately 1% V wasadded to the starting composition as analytical grade V2O5. Thestarting materials were ground under acetone to ensure fine grainsize and homogeneity, before being dried prior to the experiment.

Experiments were performed at 1.5 GPa and 1650 ◦C using anend-loaded Boyd–England-type piston-cylinder apparatus. The ex-periments employed a 12.7 mm diameter cell with an outer sleeveof pyrophyllite, an inner sleeve of SiO2 glass and a graphite heaterof outside diameter 8 mm and inside diameter 6 mm. Inside thefurnace, the SiO2 glass capsule was of 6 mm outside and 4 mminside diameters and approximately 5 mm length. The capsulehad a 1 mm thick silica glass lid and was located in the cen-tre of the assembly by 6 mm diameter machineable MgO spacerstop and bottom. Temperature in all cases was controlled using aW95Re5–W74Re26 thermocouple placed in direct contact with thecapsule.

Experiments were brought to temperature and pressure and al-lowed to equilibrate for 45 min, a time sufficient to establish Vpartitioning equilibrium (Tuff et al., 2011) before quenching. Dur-ing the experiment the SiO2 glass capsule dissolves to some extentin the silicate melt raising the SiO2 content of the latter from 47to 65 wt.% with corresponding decreases in the concentrations ofCaO, MgO, and Al2O3. The liquid Fe metal separates rapidly fromthe silicate and forms a single large ball at the bottom of thecapsule. The silicate melt quenches to a clear pale brown glass.After the experiment the capsule lid was removed and the silicateglass removed with a diamond-tipped probe and tweezers. The Femetal was recovered as a single ball. Therefore, very little crosscontamination of the two phases could have taken place. As thefinal V concentrations of metal and silicate were within a factorof 3, minor cross contamination would not have affected the ex-perimentally determined isotope fractionation factor anyway.

Using the experimental calibration of metal–silicate partition-ing of V as a function of temperature, pressure, oxygen fugacityand composition (Wood et al., 2008) our observed V partition-ing refers to an oxygen fugacity about 4 log units below theFe–FeO (IW) buffer. These conditions are relevant to the earliestphases of terrestrial accretion under which significant V parti-tioning into the core would require strongly reducing conditions(Wade and Wood, 2005). As accretion progressed the oxidationstate of the mantle implies that core segregation must have oc-curred under more oxidizing conditions, but these remained at1–2 log units below the IW buffer (Wood et al., 2008). The dom-inant oxidation state of V in silicate melts and quenched glassesat 1–4 log units below IW is V3+ (Mallmann and O’Neill, 2009;Righter et al., 2006) so our isotopic fractionation results, albeit ata single pressure and temperature, should be applicable to the his-tory of the principal phases of terrestrial core formation.

Table 2Vanadium isotope compositions of meteorites and bulk silicate Earth.

Sample Type δ51V Splitsa nb 2sd

Orgueil CI −1.9 2 4c 0.2ALH83100.202 CM2 −2.0 3 3 0.4ALH83108.40 CO3.5 −1.8 3 3 0.2Allende CV3 −1.6 5 6 0.1EET92002.15 CK4 −1.8 3 5 0.2EET96042.18 Ureilite −1.7 2 2 0.1Plainview H5 −1.7 2 3 0.3Kapoeta Howardite −1.6 3 7c 0.3Pasamonte Eucrite −1.7 2 4 0.1Johnstown Diogenite −1.7 3 11c 0.2Nakhla SNC −1.7 3 9c 0.2

Average meteorites −1.7 0.2

a Number of separate sample splits dissolved and processed through columnchemistry.

b Number of total individual mass spectrometric analyses.c Some analyses also performed at Imperial College London.

Fig. 2. Vanadium isotope compositions of meteorites obtained in this study com-pared with the bulk silicate Earth.

4. Results and comparison with bulk silicate Earth

Vanadium isotope data were obtained for five carbonaceouschondrites, one ureilite, one H5 ordinary chondrite, three achon-drites from the HED parent body and the martian meteorite Nakhla(Table 2). All samples display vanadium isotope compositions be-tween δ51V = −1.6 and −2.0� with an average value of δ51V =−1.7±0.2� (Fig. 2). We also report results for three metal–silicateequilibration experiments (Table 3, Fig. 3). These data show that,within error, there is no appreciable V isotope fractionation be-tween liquid metal and silicate melt.

The value obtained for extraterrestrial material is significantlydifferent from the bulk silicate Earth (BSE), which is estimatedat δ51V = −0.7 ± 0.2� (Prytulak et al., 2013). Hence, BSE and

170 S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175

Table 3Results of high pressure experiments.

Experiment [V]metal(ppm)

[V]sil(ppm)

D δ51Vmetal 2sd n δ51Vsil 2sd n �51Vmet-sil

0906 3500 12 200 0.29 −1.21 0.10 3 −1.16 0.09 4 −0.050914 3900 10 200 0.38 −1.21 0.11 5 −1.41 0.12 4 0.200917 3100 14 200 0.22 −1.34 0.08 5 −1.23 0.12 5 −0.11

Fig. 3. Vanadium isotope fractionation (�51V) between metal and silicate portionsin three experimental charges. Very little isotope fractionation is registered in theseexperiments, which is consistent with theoretical predictions for Fe and Cr, tran-sition metals of similar mass. If Earth was chondritic for vanadium isotopes thenit is required that �51V > 1.4 (see Fig. 4) in order to explain the silicate Earth’svanadium isotope deficit.

chondrites are different by 1� (Fig. 2), which is equivalent to∼14 standard deviations. A recent study of terrestrial basalts andperidotites provides an extensive discussion of the V isotope com-position of BSE (Prytulak et al., 2013). Briefly, mantle derivedbasalts, which encompass mid-ocean ridge basalts from Atlantic,Pacific and Indian Ocean basins and tholeiitic basalts from Ice-land, are invariant with an average value of δ51V = −0.9 ± 0.2�(2sd). Peridotites (both abyssal and orogenic) show more scat-ter, but fall within the same range of values as mantle derivedbasalts. The total range of vanadium isotope compositions recordedfor peridotites (both highly depleted and fertile) and basalts isδ51V = −1.2 to −0.6�. Fresh, fertile peridotites have a relativelyuniform value of δ51V = −0.7 ± 0.2� and since >99% of vana-dium in the BSE is situated in the mantle (Prytulak et al., 2013)fertile peridotites are used as the best estimate for BSE. The ex-isting basalt and peridotite data argue strongly that the BSE doesnot have a V isotope composition as light as that measured inmeteorites (Table 2). In the following section we discuss poten-tial reasons for this significant difference between BSE and mete-orites.

5. Possible origins of vanadium isotope variation in the SolarSystem

Vanadium has only two isotopes, so it is not possible to distin-guish between mass dependent and independent isotope effects.The heavy V isotope composition of BSE can potentially be ex-plained by a variety of processes. These include:

(A) Terrestrial core formation (metal–silicate equilibrium fraction-ation);

(B) Volatile element depletion or impact erosion;(C) Stable isotope fractionation during igneous processing;(D) Stable isotope fractionation during parent body metamor-

phism;(E) Cosmic irradiation of meteorites before delivery on Earth;(F) Heterogeneous distribution of nucleosynthetically anomalous

material;(G) High-energy irradiation of the solar nebula before or during

formation of the first nebular condensates such as CAIs.

Below we outline each of these processes together with anassessment of the plausibility or otherwise that they were respon-sible for the observed V isotopic offset between Earth and mete-orites.

5.1. Terrestrial core formation

Experimental studies have shown that vanadium should par-tition, to some extent, into the core (Chabot and Agee, 2003;Wade and Wood, 2005; Wood et al., 2008) and comparisons be-tween the vanadium content of BSE and that expected based onchondrite meteorites show that the core may contain 40–50% ofEarth’s vanadium (McDonough and Sun, 1995; O’Neill, 1991). Giventhat BSE is characterized by δ51V ∼ −0.7� and chondrites (e.g. inthis calculation bulk Earth) display δ51V ∼ −1.7�, we can calcu-late the amount of isotope fractionation between the Earth’s coreand mantle required to explain the isotopic difference between BSEand meteorites via core formation. In Fig. 4 we use the isotopicmass balance equation:

δ51VBE = FV,CORE × δ51VCORE + FV,BSE × δ51VBSE (1)

where δ51VBE, δ51VCORE and δ51VBSE are the vanadium isotopecompositions of bulk Earth (assumed to be chondritic), the coreand BSE, respectively. FV,CORE and FV,BSE are the fractions of vana-dium residing in the core and silicate portions of the Earth andadd up to 1. The minimum fractionation factor required to accountfor the difference between BSE and chondrites by core formationis �51V = 2 (Fig. 4), which is very large considering the hightemperatures of core formation as well as the small relative massdifference between the two vanadium isotopes of ∼2%. We haveinvestigated the feasibility of this process by measuring the vana-dium isotope fractionation in experiments where liquid metal andsilicate were equilibrated. Within the analytical error of between0.1 and 0.2� we find no significant vanadium isotopic fractiona-tion at equilibrium between liquid metal and liquid silicate (Fig. 3).This is a reasonable result given that the isotopes of Cr, Fe andNi, elements of similar mass, both via experimental and theo-retical means have been shown to fractionate by at most 0.1�per amu at the conditions of core formation (Lazar et al., 2012;Moynier et al., 2011; Poitrasson et al., 2009; Polyakov, 2009). Thus,there is currently no evidence to suggest that vanadium isotopeswere fractionated significantly when vanadium partitioned into thecore.

5.2. Volatile element depletion or impact erosion

The Earth is depleted in volatile elements and even though Vis relatively refractory with a half mass condensation temperature

S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175 171

Fig. 4. Calculated vanadium isotope composition of the Earth’s core as a function offraction total Earth V in the core. If it is assumed that the bulk Earth is chondritic(δ51V = −1.7) then isotope fractionation between the core and mantle (bulk silicateEarth) can be calculated using either a batch (a) or a fractional (b) core formationmodel. It can be seen that for the batch model (a), the isotope fractionation (�51V)between metal and silicate must be between 2 and 2.5 δ51V-units in order for Earthto be chondritic. Fractional core formation (b) yields lower values between �51V =1.4–1.9 δ51V-units.

of 1427 K (Lodders, 2003), it is possible that a fraction of vana-dium was lost to space during some type of transient heatingevent either before, during or after the Moon-forming giant im-pact. Volatile loss may be associated with kinetic isotope frac-tionation and, if it is assumed that the volatile loss event onlyaffected the 1AU region, then this process could account forthe heavy V isotope signature of Earth compared with essen-tially uniform V isotopes in Mars and the Asteroid Belt (Fig. 2).However, volatile loss of vanadium from Earth should also af-fect the stable isotope compositions of a number of elementswith similar volatility. Specifically, Sr, Ca and Ni all have con-densation temperatures within 80 K of vanadium (Lodders, 2003)and would be expected to display large kinetic effects, becausetheir isotopes have larger relative mass differences than vanadium.However, recent high-precision measurements of Earth, chondritesand achondrites reveal that Earth is isotopically indistinguishablefrom meteorites for these three elements (Charlier et al., 2012;Moynier et al., 2007, 2010; Simon and Depaolo, 2010; Steele etal., 2011), which leaves a volatile loss explanation for the vana-dium isotope difference between bulk Earth and meteorites highlyunlikely.

Impact erosion is the process whereby portions of the al-ready differentiated Earth are lost to space via meteorite impacts

(O’Neill and Palme, 2008). This process supposedly removes pri-marily crustal material (i.e. the outermost few kilometers of thesilicate Earth) thereby depleting bulk Earth in a number of incom-patible elements that are enriched in this reservoir (Campbell andO’Neill, 2012). While this model may account for several chemicalfeatures of Earth, it is highly unlikely to have affected vanadiumisotopes, because vanadium is only mildly incompatible. Presently,less than 1% of the V in BSE resides in the crust and thus even ifan entire crustal reservoir had been removed via collisional erosionthere would be negligible effect on vanadium isotopes.

5.3. Stable isotope fractionation during igneous or parent bodyprocessing

Vanadium isotope fractionation exists at igneous temperaturesas documented by significant vanadium isotope variability in ter-restrial igneous rocks (Prytulak et al., 2011, 2013). However, sev-eral observations strongly indicate that the estimate of BSE vana-dium isotope composition is not in error to the extent that thesilicate Earth is similar to meteorites.

Firstly, vanadium is only slightly incompatible during mantlemelting (Canil, 1999; Canil and Fedortchouk, 2000; Lee et al.,2003), which means that >99% of vanadium in the silicate Earthresides in the mantle. Thus, more evolved igneous rocks like an-desite and granodiorite, which are characterized by heavier V iso-tope compositions (δ51V = −0.4 to −0.6�; Prytulak et al., 2011)do not have a significant influence on the average BSE. Therefore,only mantle derived basalts or peridotites should be consideredgood analogues for the V isotope composition of BSE. Moreover,the limited data for V isotope compositions of more evolved ig-neous materials (Prytulak et al., 2011) are heavier than BSE es-timates. Hence adding evolved lithologies to our current BSE es-timate would increase the difference between terrestrial and ex-traterrestrial materials.

Secondly, the entire range of terrestrial igneous rocks so far in-vestigated (>60 samples), which include depleted and fertile peri-dotites, mid ocean ridge basalts, primitive basalts from Iceland andHawaii, and one granodiorite, span values from δ51V = −0.3 to−1.1, with the vast majority of mantle derived samples (i.e. peri-dotites and basalts) falling between δ51V = −0.7 and −0.9. Noterrestrial sample measured thus far even approaches the isotopecomposition of the heaviest meteorite (the Howardite Kapoeta,δ51V = −1.6 ± 0.3, Table 2).

Thirdly, a few highly depleted peridotites (Al2O3 < 0.6%) thathave very low V concentrations display slightly lighter V iso-tope compositions (δ51V ∼ −1.1) than basalts and fertile peri-dotites, which may imply vanadium isotope fractionation dur-ing mantle melting (Prytulak et al., 2013). However, the averageprimitive mantle is composed of lherzolitic material with Al2O3∼4.0–4.5 wt.% (McDonough and Sun, 1995; Salters and Stracke,2004; Workman and Hart, 2005), so these ultra-depleted peri-dotites are unlikely to represent the average mantle vanadium iso-tope composition. However, even the lightest, depleted peridotitesagain are not as light as the heaviest meteorite measurements.

Combined with the evidence for negligible V isotope fraction-ation during core formation (Fig. 3), it is therefore evident thatbulk Earth is heavier than all extraterrestrial material so far in-vestigated by at least 0.9� (compared with Kapoeta). It is alsounlikely that other stable isotope fractionation mechanisms likecondensation/evaporation and parent body alteration processes canaccount for the difference between BSE and meteorites. Vanadiumis refractory and thus not prone to thermal metamorphism onmeteorite parent bodies. Similarly, aqueous alteration does not sig-nificantly affect vanadium or its isotopes (Prytulak et al., 2013;Teagle et al., 1996). In any event, if secondary stable vanadium iso-tope fractionation processes were the cause of the BSE–meteorite

172 S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175

Fig. 5. Vanadium isotope composition plotted versus cosmic ray exposure ages forseven meteorites: Orgueil (Mazor et al., 1970), Allende (Eugster et al., 2007), Plain-view (Graf and Marti, 1995), Johnstown (Welten et al., 1997), Pasamonte (Eugsterand Michel, 1995), Kapoeta (Eugster and Michel, 1995), Nakhla (Eugster et al., 1997).Also shown are the isotope composition of the bulk silicate Earth and the hypothet-ical general direction that would be expected if vanadium isotopes were affectedcosmic ray spallation.

offset then it would be extraordinarily fortuitous that all mete-orites irrespective of volatile depletion and degree of parent bodymetamorphism display similar δ51V.

5.4. Cosmic irradiation of meteorites before delivery to Earth

An obvious difference between meteorites and terrestrial sam-ples is that meteorites have spent a significant amount timein space before colliding with Earth. Since it is possible that50V is produced by interaction with high-energy cosmic particles(Gounelle et al., 2001; Lee et al., 1998) the lighter δ51V valuesrecorded for the meteorites could simply be explained by cosmicray spallation processes that took place in the time between parentbody breakup and arrival on Earth. However, this conclusion is notsupported by the variable cosmic ray exposure ages of between 4and 22 Ma recorded for some of the samples studied here (Eugsteret al., 2007; Eugster and Michel, 1995; Graf and Marti, 1995;Mazor et al., 1970; Welten et al., 1997). These variations in ex-posure ages would be expected to cause an appreciable spreadin vanadium isotope compositions if recent cosmic ray exposurewere the cause of the light vanadium isotope signatures in themeteorites (Fig. 5). Therefore, cosmic ray exposure is an unlikelyexplanation for the V isotope composition of the studied mete-orites.

5.5. Nucleosynthetic heterogeneity of vanadium in the Solar System

Many elements display isotopic anomalies on the bulk me-teorite scale that are clearly related to incomplete mixing ofmaterial synthesized in specific stellar environments (Andreasenand Sharma, 2006; Brandon et al., 2005; Burkhardt et al., 2011;Carlson et al., 2007; Dauphas et al., 2010; Trinquier et al., 2009;Zhang et al., 2012). In theory, such processes could explain theheavy V isotope composition of Earth. In order to assess thispossibility, knowledge of the relative nucleosynthetic productionrates of 50V and 51V is required. While 51V most likely is pre-dominantly produced by nuclear statistical equilibrium in super-novae (Cameron, 1973; Woosley et al., 2002), a number of stellarprocesses appear capable of explaining the Solar System budgetof 50V. These include carbon, neon and oxygen burning in mas-sive stars (Woosley et al., 2002) as well as spallogenic production

during galactic cosmic ray exposure in the interstellar medium(Audouze, 1970). Given these uncertainties, it is currently diffi-cult to predict specific V isotopic signatures associated with stellarsources and thus infer whether Earth might be depleted or en-riched in nucleosynthetically synthesized V compared with me-teorites. In either case, however, an explanation of our data vianucleosynthetic anomalies must involve the Earth being uniformlyenriched or depleted in a nucleosynthetic component comparedwith carbonaceous chondrites, Mars and Vesta.

This requirement is in stark contrast to every other isotopesystem for which bulk meteorite nucleosynthetic variation hasbeen identified (Brandon et al., 2005; Burkhardt et al., 2011;Carlson et al., 2007; Regelous et al., 2008; Trinquier et al., 2009;Zhang et al., 2012) where carbonaceous chondrites ubiquitouslydisplay more isotopic variation and different values compared withordinary chondrites and achondrites. This relationship is true fornucleosynthetic anomalies in Ti, Cr, Ni, Ru, Mo, Ba, Sm, Nd andOs isotopes. Thus, since V isotopes are similar in carbonaceouschondrites and achondrites it is highly unlikely that the precur-sor material, which caused the nucleosynthetic anomalies in theabove-listed elements, would also be characterized by fractionatedV isotope compositions. We cannot at present exclude the possi-bility that some V was supplied to the Solar System from an asyet unidentified stellar source that did not carry large concentra-tions or isotopic anomalies of the above listed elements, whichthen caused the V isotope variation we observe between Earth andmeteorites. However, given these constraints on the stellar sourceof V, it is difficult to reconcile the current V isotope data set interms of nucleosynthetic anomalies.

5.6. Production of 50V during irradiation in the early Solar System

As outlined in the introduction, vanadium isotope anomaliesare produced during high-energy irradiation. Thus, in principle,early Solar System irradiation could explain the ∼1� differencebetween Earth and the Asteroid Belt and Mars (Fig. 2). Quali-tatively the isotopic difference translates into the Asteroid Beltand Mars being enriched in material that was irradiated by high-energy particles. It is important to note that the difference be-tween BSE and meteorites cannot be exclusively explained bythe presence of CAIs in the Asteroid Belt and Mars that areenriched in 50V via early Solar System irradiation because car-bonaceous chondrites exhibit highly variable amounts of CAIs andare indistinguishable in terms of V isotopes. Thus, although CAIsmay possess light V isotope compositions (Gounelle et al., 2001;Lee et al., 1998), CAIs by themselves do not appear to be the pri-mary cause of isotopic variations between Earth and meteorites.

The qualitative assessment that the Asteroid Belt and Mars areenriched in irradiated material compared with Earth may seemcounter intuitive because the Earth is situated closer to the Sun,which was the source of irradiation. Thus, one might expect thatEarth-forming materials would experience higher degrees of irra-diation. However, the observed deficit of irradiated vanadium inEarth could be consistent with a model, in which material be-comes irradiated by high-energy particles from solar flares whichwas subsequently ejected to the Mars and Asteroid Belt regions viamagnetocentrifugal outflow (Coffey et al., 2007; Reipurth and Bally,2001). The ejected material could, for example, be dominated bymicron sized or smaller particles rich in refractory elements, whichthen mix with matter already present at the radial distance ofthe Asteroid Belt and Mars. If we assume that these particles con-tain approximately the same vanadium concentration (∼500 ppm)(Sylvester et al., 1992) and δ51V ∼ −4.5 (Gounelle et al., 2001;Lee et al., 1998), as inferred for CAIs irradiated in an X-wind sce-nario, we can calculate that the Asteroid Belt and Mars shouldcontain about 2% more by mass of this component than Earth.

S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175 173

Other elements are known to be sensitive to energetic parti-cle reactions, in particular the spallogenic elements lithium (Li)and boron (B) as well as lanthanum (La), which due to the lowabundance of 138La (∼0.09%) is also sensitive to irradiation. The Visotope data can be evaluated in the context of these three ele-ments and their isotopic distributions in the Solar System. Seitz etal. (2007) discussed the isotopic effects on Li and B from spallationand suggested that both should become enriched in the light iso-tope (6Li and 10B). These calculations included direct productionof the stable Li and B isotopes as well as contributions from ra-dioactive decay of spallogenic 7Be (to 7Li) and 10Be (to 10B). Thesecalculations, therefore, indicate that enrichments in 50V should beaccompanied by correlated enrichments in 6Li and 10B. Based onthis reasoning, Earth should be characterized by isotopically lightLi and B isotope compositions compared with carbonaceous chon-drites, Vesta and Mars, which is not observed (Magna et al., 2006;Seitz et al., 2007; Zhai et al., 1996). However, since both Li andB are relatively volatile (50% condensation temperatures of about1140 K and 900 K, respectively; Lodders, 2003) it is unlikely thatthey would be transported with the refractory material whichmay have carried isotopically light V to the Mars/Asteroid Beltregion. Beryllium has approximately the same condensation tem-perature as V and thus spallogenic 7Be and 10Be (that woulddecay to 7Li and 10B, respectively) could be delivered with theisotopically light V. However, if the suggested production rates of7Be/9Be = 6 × 10−3 and 10Be/9Be = 1 × 10−3 (Chaussidon etal., 2006) are combined with a maximum CAI Be concentrationof 0.09 ppm (Chaussidon et al., 2006) and Solar System (i.e. CIchondrites) [Li] = 1.5 ppm and [B] = 0.9 ppm (McDonough andSun, 1995) the calculated effects on δ7Li and δ11B in bulk mete-orites are less than 0.5�, which would not be detectable withinthe isotopic variation observed for Earth and meteorites especiallygiven that the uncertainties on the Li and B isotopic composi-tions of Earth are larger than 0.5� (Chaussidon and Marty, 1995;Elliott et al., 2004).

Lanthanum consists of two isotopes (masses 138 and 139)and has a condensation temperature of ∼1570 K (Lodders, 2003),which is higher than V (∼1430 K). Thus, the refractory materialthat may have transported significant amounts of V would be sim-ilarly enriched in La. Since the 138La isotope anomalies generatedby irradiation are of the same magnitude as V isotopes (i.e. 2–7�;Gounelle et al., 2001; Lee et al., 1998) it is predicted that La andV isotope compositions of bulk meteorites would display similarvariation (on the order of 1�) as well as a significant correla-tion if their isotope compositions were largely due to incorpora-tion of irradiated material. Lanthanum isotope data are scant andonly a few analyses for CAIs and bulk meteorites that have rel-atively large errors of close to ±1� exist (Shen and Lee, 2003;Shen et al., 1994). However, there is a suggestion that La isotopecompositions of carbonaceous chondrites are enriched in 138La byabout 1� compared with Earth (Shen et al., 1994); consistent withthe expected shift if V and La isotopes were affected by irradiation.However, given the large uncertainty on the small number of Laisotope data, it is premature to discuss any significant correlationbetween V and La isotopes.

Although the above reasoning appears to favor an irradiationorigin for the Earth–meteorite V isotopic offset, this explanationhas significant problems. For example, it is unclear how irradi-ated material was transported to Mars and the Asteroid Belt inequal proportions, while the 1AU region was left depleted. In ad-dition, the strikingly uniform V isotope compositions of the dif-ferent types of meteorites (carbonaceous and ordinary chondrites,achondrites) poses a significant problem in that it requires thatthe irradiated refractory disk material in the region of Mars andthe Asteroid Belt was already well mixed before being incorpo-rated into planetary bodies. This requirement is difficult to recon-

cile with the compositional variation of carbonaceous chondrites,which argues against a well-mixed nebula at the time of their for-mation. Therefore, more data are needed before the cause of thesignificant vanadium isotope difference between the Earth and me-teorites can be determined with confidence.

6. Implications for the chemical composition of Earth

Even though it is presently difficult to offer a unique explana-tion for the vanadium isotopic difference between BSE and me-teorites, it is highly unlikely that bulk Earth has identical δ51Vto chondrites (see Sections 5.1, 5.3 and 5.4). This indicates thatthe chemical composition of Earth cannot entirely be accountedfor by mixing chondritic meteorites in various proportions (e.g.McDonough and Sun, 1995; Ringwood, 1966). This conclusion is inagreement with several studies that also argue for a non-chondriticEarth (e.g. Burkhardt et al., 2011; Campbell and O’Neill, 2012;Drake and Righter, 2002; O’Neill and Palme, 2008). It should bestressed that the V isotopic evidence presented here may be com-patible with the majority of Earth being composed of chondriticmeteorites, such that only a small fraction of isotopically fraction-ated material need be added to account for the terrestrial V isotopeoffset.

Since there are some notably absent classes of meteorites (e.g.enstatite chondrites) in our V isotope data, we cannot exclude thepossibility that Earth is composed of a mixture of other types ofchondritic material. However, if the meteorites investigated here,namely carbonaceous chondrites, ureilites and H chondrites, to-gether comprise a major portion of Earth (e.g. Fitoussi and Bour-don, 2012; Warren, 2011) then there must be very large V isotopevariations present in the remaining classes of chondrites (L, LL, EL,EH) to reconcile the Earth–meteorite δ51V difference.

Acknowledgements

The Natural History Museum of London is thanked for supply-ing samples. Barry Coles (Imperial) and Nick Belshaw (Oxford) arethanked for instrument support. We would also like to thank TimElliott for his editorial handling and two anonymous reviewers,whose comments improved the manuscript. This study was sup-ported by NERC fellowships to S.G.N. and J.P. (NE/H01313X/1 (andX/2)) and grants from Petrobras, STFC and ERC to A.N.H. and B.J.W.Funds from the Penzance endowed fund in support of assistant sci-entists (S.G.N.) are also acknowledged.

References

Andreasen, R., Sharma, M., 2006. Solar nebula heterogeneity in p-process samariumand neodynium isotope. Science 314, 806–809.

Audouze, J., 1970. Some nucleosynthetic effects of energetic proton fluxes on stellarsurfaces and on Solar System. Astron. Astrophys. 8, 436–457.

Balsiger, H., Geiss, J., Lipschutz, M.E., 1969. Vanadium isotopic composition in mete-oritic and terrestrial matter. Earth Planet. Sci. Lett. 6, 117–122.

Balsiger, H., Mendia, M.D., Pelly, I.Z., Lipschutz, M.E., 1976. Vanadium isotopic com-position and contents in gas-rich meteorites. Earth Planet. Sci. Lett. 28, 379–384.

Brandon, A.D., Humayun, M., Puchtel, I.S., Leya, I., Zolensky, M., 2005. Osmium iso-tope evidence for an s-process carrier in primitive chondrites. Science 309,1233–1236.

Burkhardt, C., Kleine, T., Oberli, F., Pack, A., Bourdon, B., Wieler, R., 2011. Molybde-num isotope anomalies in meteorites: Constraints on solar nebula evolution andorigin of the Earth. Earth Planet. Sci. Lett. 312, 390–400.

Burnett, D.S., Fowler, W.A., Hoyle, F., 1965. Nucleosynthesis in the early history ofthe solar system. Geochim. Cosmochim. Acta 29, 1209–1241.

Cameron, A.G.W., 1973. Abundances of elements in Solar System. Space Sci. Rev. 15,121–146.

Campbell, I.H., O’Neill, H.S.C., 2012. Evidence against a chondritic Earth. Nature 483,553–558.

Canil, D., 1999. Vanadium partitioning between orthopyroxene, spinel and sili-cate melt and the redox states of mantle source regions for primary magmas.Geochim. Cosmochim. Acta 63, 557–572.

174 S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175

Canil, D., Fedortchouk, Y., 2000. Clinopyroxene–liquid partitioning for vanadium andthe oxygen fugacity during formation of cratonic and oceanic mantle litho-sphere. J. Geophys. Res., Solid Earth 105, 26003–26016.

Carlson, R.W., Boyet, M., Horan, M., 2007. Chondrite barium, neodymium, andsamarium isotopic heterogeneity and early earth differentiation. Science 316,1175–1178.

Chabot, N.L., Agee, C.B., 2003. Core formation in the Earth and Moon: New ex-perimental constraints from V, Cr, and Mn. Geochim. Cosmochim. Acta 67,2077–2091.

Charlier, B.L.A., Nowell, G.M., Parkinson, I.J., Kelley, S.P., Pearson, D.G., Burton, K.W.,2012. High temperature strontium stable isotope behaviour in the early solarsystem and planetary bodies. Earth Planet. Sci. Lett. 329–330, 31–40.

Chaussidon, M., Marty, B., 1995. Primitive boron isotope composition of the mantle.Science 269, 383–386.

Chaussidon, M., Robert, F., McKeegan, K.D., 2006. Li and B isotopic variations inan Allende CAI: Evidence for the in situ decay of short-lived Be-10 and forthe possible presence of the short-lived nuclide Be-7 in the early solar system.Geochim. Cosmochim. Acta 70, 224–245.

Coffey, D., Bacciotti, F., Ray, T.P., Eisloeffel, J., Woitas, J., 2007. Further indications ofjet rotation in new ultraviolet and optical hubble space telescope STIS spectra.Astrophys. J. 663, 350–364.

Dauphas, N., Remusat, L., Chen, J.H., Roskosz, M., Papanastassiou, D.A., Stodolna, J.,Guan, Y., Ma, C., Eiler, J.M., 2010. Neutron-rich chromium isotope anomalies insupernova nanoparticles. Astrophys. J. 720, 1577.

Desch, S.J., Morris, M.A., Connolly Jr., H.C., Boss, A.P., 2010. A critical examinationof the X-wind model for chondrule and calcium-rich, aluminum-rich inclusionformation and radionuclide production. Astrophys. J. 725, 692–711.

Drake, M., Righter, K., 2002. Determining the composition of the Earth. Nature 416,39–44.

Elliott, T., Jeffcoate, A., Bouman, C., 2004. The terrestrial Li isotope cycle: light-weight constraints on mantle convection. Earth Planet. Sci. Lett. 220, 231–245.

Eugster, O., Michel, T., 1995. Common asteroid break-up events of eucrites, diogen-ites and howardites and cosmic-ray production-rates for noble-gases in achon-drites. Geochim. Cosmochim. Acta 59, 177–199.

Eugster, O., Weigel, A., Polnau, E., 1997. Ejection times of Martian meteorites.Geochim. Cosmochim. Acta 61, 2749–2757.

Eugster, O., Lorenzetti, S., Kraehenbuehl, U., Marti, K., 2007. Comparison of cosmic-ray exposure ages and trapped noble gases in chondrule and matrix samplesof ordinary, enstatite, and carbonaceous chondrites. Meteorit. Planet. Sci. 42,1351–1371.

Feigelson, E.D., 2010. X-ray insights into star and planet formation. Proc. Natl. Acad.Sci. USA 107, 7153–7157.

Fitoussi, C., Bourdon, B., 2012. Silicon isotope evidence against an enstatite chondriteearth. Science 335, 1477–1480.

Fowler, W.A., Greenstein, J.L., Hoyle, F., 1962. Nucleosynthesis during the early his-tory of the solar system. Geophys. J. R. Astron. Soc. 6, 148–220.

Gounelle, M., Shu, F.H., Shang, H., Glassgold, A.E., Rehm, K.E., Lee, T., 2001. Extinctradioactivities and protosolar cosmic rays: Self-shielding and light elements. As-trophys. J. 548, 1051–1070.

Graf, T., Marti, K., 1995. Collisional history of H-chondrites. J. Geophys. Res., Plan-ets 100, 21247–21263.

Jacobsen, B., Matzel, J., Hutcheon, I.D., Krot, A.N., Yin, Q.-Z., Nagashima, K., Ramon,E.C., Weber, P.K., Ishii, H.A., Ciesla, F.J., 2011. Formation of the short-lived ra-dionuclide (36)Cl in the protoplanetary disk during late-stage irradiation of avolatile-rich reservoir. Astrophys. J. Lett. 731.

Lazar, C., Young, E.D., Manning, C.E., 2012. Experimental determination of equilib-rium nickel isotope fractionation between metal and silicate from 500 degrees Cto 950 degrees C. Geochim. Cosmochim. Acta 86, 276–295.

Lee, T., Shu, F.H., Shang, H., Glassgold, A.E., Rehm, K.E., 1998. Protostellar cosmic raysand extinct radioactivities in meteorites. Astrophys. J. 506, 898–912.

Lee, C.T.A., Brandon, A.D., Norman, M., 2003. Vanadium in peridotites as a proxyfor paleo-fO(2) during partial melting: Prospects, limitations, and implications.Geochim. Cosmochim. Acta 67, 3045–3064.

Lodders, K., 2003. Solar system abundances and condensation temperatures of theelements. Astrophys. J. 591, 1220–1247.

Magna, T., Wiechert, U., Halliday, A.N., 2006. New constraints on the lithium isotopecompositions of the Moon and terrestrial planets. Earth Planet. Sci. Lett. 243,336–353.

Mallmann, G., O’Neill, H.S.C., 2009. The crystal/melt partitioning of V during mantlemelting as a function of oxygen fugacity compared with some other elements(Al, P, Ca, Sc, Ti, Cr, Fe, Ga, Y, Zr and Nb). J. Petrol. 50, 1765–1794.

Mazor, E., Heymann, D., Anders, E., 1970. Noble gases in carbonaceous chondrites.Geochim. Cosmochim. Acta 34, 781–824.

McDonough, W.F., Sun, S.-s., 1995. The composition of the Earth. Chem. Geol. 120,223–253.

McKeegan, K.D., Chaussidon, M., Robert, F., 2000. Incorporation of short-lived Be-10in a calcium–aluminum-rich inclusion from the Allende meteorite. Science 289,1334–1337.

Moynier, F., Blichert-Toft, J., Telouk, P., Luck, J.M., Albarede, F., 2007. Comparative sta-ble isotope geochemistry of Ni, Cu, Zn, and Fe in chondrites and iron meteorites.Geochim. Cosmochim. Acta 71, 4365–4379.

Moynier, F., Agranier, A., Hezel, D.C., Bouvier, A., 2010. Sr stable isotope composi-tion of Earth, the Moon, Mars, Vesta and meteorites. Earth Planet. Sci. Lett. 300,359–366.

Moynier, F., Yin, Q.Z., Schauble, E., 2011. Isotopic evidence of Cr partitioning intoEarth’s core. Science 331, 1417–1420.

Nielsen, S.G., Prytulak, J., Halliday, A.N., 2011. Determination of precise and accurate51V–50V isotope ratios by MC-ICP-MS, Part 1: Chemical separation of vanadiumand mass spectrometric protocols. Geostand. Geoanal. Res. 35, 293–306.

O’Neill, H.S.C., 1991. The origin of the Moon and the early history of the Earth –A chemical model. Part 2: The Earth. Geochim. Cosmochim. Acta 55, 1159–1172.

O’Neill, H.S.C., Palme, H., 2008. Collisional erosion and the non-chondritic compo-sition of the terrestrial planets. Philos. Trans. R. Soc., Math. Phys. Eng. Sci. 366,4205–4238.

Pelly, I.Z., Lipschutz, M.E., Balsiger, H., 1970. Vanadium isotopic composition andcontents in chondrites. Geochim. Cosmochim. Acta 34, 1033–1036.

Poitrasson, F., Roskosz, M., Corgne, A., 2009. No iron isotope fractionation betweenmolten alloys and silicate melt to 2000 degrees C and 7.7 GPa: Experimental ev-idence and implications for planetary differentiation and accretion. Earth Planet.Sci. Lett. 278, 376–385.

Polyakov, V.B., 2009. Equilibrium iron isotope fractionation at core–mantle boundaryconditions. Science 323, 912–914.

Presnall, D.C., Dixon, S.A., Dixon, J.R., O’Donnell, T.H., Brenner, N.L., Schrock, R.L.,Dycus, D.W., 1978. Liquidus phase relations on the join diopside–forsterite–anorthite from 1 atm to 20 kbar: Their bearing on the generation and crys-tallization of basaltic magma. Contrib. Mineral. Petrol. 66, 203–220.

Prytulak, J., Nielsen, S.G., Halliday, A.N., 2011. Determination of precise and accurate51V/50V isotope ratios by multi-collector ICP-MS, Part 2: Isotopic composition ofsix reference materials plus the Allende chondrite and verification tests. Geo-stand. Geoanal. Res. 35, 307–318.

Prytulak, J., Nielsen, S.G., Ionov, D.A., Halliday, A.N., Harvey, J., Kelley, K.A., Niu, Y.,Peate, D.W., Shimizu, K., Sims, K.W.W., 2013. The stable vanadium isotope com-position of the mantle and mafic lavas. Earth Planet. Sci. Lett. 365, 177–189.

Regelous, M., Elliott, T., Coath, C.D., 2008. Nickel isotope heterogeneity in the earlySolar System. Earth Planet. Sci. Lett. 272, 330–338.

Reipurth, B., Bally, J., 2001. Herbig–Haro flows: Probes of early stellar evolution.Annu. Rev. Astron. Astrophys. 39, 403–455.

Righter, K., Sutton, S.R., Newville, M., Lei, L., Schwandt, C.S., Uchida, H., Lavina, B.,Downs, R.T., 2006. An experimental study of the oxidation state of vanadium inspinel and basaltic melt with implications for the origin of planetary basalt. Am.Mineral. 91, 1643–1656.

Ringwood, A.E., 1966. The chemical composition and origin of the Earth. In: Hurley,P.M. (Ed.), Advances in Earth Sciences. MIT Press, Cambridge, pp. 287–356.

Salters, V.J.M., Stracke, A., 2004. Composition of the depleted mantle. Geochem. Geo-phys. Geosyst. 5.

Seitz, H.M., Brey, G.P., Zipfel, J., Ott, U., Weyer, S., Durali, S., Weinbruch, S., 2007.Lithium isotope composition of ordinary and carbonaceous chondrites, anddifferentiated planetary bodies: Bulk solar system and solar reservoirs. EarthPlanet. Sci. Lett. 260, 582–596.

Shen, J.J.S., Lee, T., 2003. La-138 anomaly in the early solar system. Astrophys. J. 596,L109–L112.

Shen, J.J.S., Lee, T., Chang, C.T., 1994. Lanthanum isotopic composition of meteoriticand terrestrial matter. Geochim. Cosmochim. Acta 58, 1499–1506.

Shu, F.H., Shang, H., Lee, T., 1996. Toward an astrophysical theory of chondrites.Science 271, 1545–1552.

Shu, F.H., Shang, H., Glassgold, A.E., Lee, T., 1997. X-rays and fluctuating x-windsfrom protostars. Science 277, 1475–1479.

Shu, F.H., Shang, H., Gounelle, M., Glassgold, A.E., Lee, T., 2001. The origin of chon-drules and refractory inclusions in chondritic meteorites. Astrophys. J. 548,1029–1050.

Simon, J.I., Depaolo, D.J., 2010. Stable calcium isotopic composition of meteoritesand rocky planets. Earth Planet. Sci. Lett. 289, 457–466.

Steele, R.C.J., Elliott, T., Coath, C.D., Regelous, M., 2011. Confirmation of mass-independent Ni isotopic variability in iron meteorites. Geochim. Cosmochim.Acta 75, 7906–7925.

Sylvester, P.J., Grossman, L., MacPherson, G.J., 1992. Refractory inclusions withunusual chemical compositions from the Vigarano carbonaceous chondrite.Geochim. Cosmochim. Acta 56, 1343–1363.

Teagle, D.A.H., Alt, J.C., Bach, W., Halliday, A.N., Erzinger, J., 1996. Alteration of upperocean crust in a ridge-flank hydrothermal upflow zone: Mineral, chemical, andisotopic constraints from hole 896A. Proc. Ocean Drill. Program Sci. Results 148,119–150.

Trinquier, A., Elliott, T., Ulfbeck, D., Coath, C., Krot, A.N., Bizzarro, M., 2009. Originof nucleosynthetic isotope heterogeneity in the solar protoplanetary disk. Sci-ence 324, 374–376.

Tuff, J., Wood, B.J., Wade, J., 2011. The effect of Si on metal–silicate partitioningof siderophile elements and implications for the conditions of core formation.Geochim. Cosmochim. Acta 75, 673–690.

Wade, J., Wood, B.J., 2005. Core formation and the oxidation state of the Earth. EarthPlanet. Sci. Lett. 236, 78–95.

S.G. Nielsen et al. / Earth and Planetary Science Letters 389 (2014) 167–175 175

Warren, P.H., 2011. Stable-isotopic anomalies and the accretionary assemblage of theEarth and Mars: A subordinate role for carbonaceous chondrites. Earth Planet.Sci. Lett. 311, 93–100.

Welten, K.C., Lindner, L., van der Borg, K., Loeken, T., Scherer, P., Schultz, L., 1997.Cosmic-ray exposure ages of diogenites and the recent collisional history of thehowardite, eucrite and diogenite parent body/bodies. Meteorit. Planet. Sci. 32,891–902.

Wood, J.A., 2004. Formation of chondritic refractory inclusions: the astrophysicalsettings. Geochim. Cosmochim. Acta 68, 4007–4021.

Wood, B.J., Wade, J., Kilburn, M.R., 2008. Core formation and the oxidation state

of the Earth: Additional constraints from Nb, V and Cr partitioning. Geochim.Cosmochim. Acta 72, 1415–1426.

Woosley, S.E., Heger, A., Weaver, T.A., 2002. The evolution and explosion of massivestars. Rev. Mod. Phys. 74, 1015–1071.

Workman, R.K., Hart, S.R., 2005. Major and trace element composition of the de-pleted MORB mantle (DMM). Earth Planet. Sci. Lett. 231, 53–72.

Zhai, M.Z., Nakamura, E., Shaw, D.M., Nakano, T., 1996. Boron isotope ratios in me-teorites and lunar rocks. Geochim. Cosmochim. Acta 60, 4877–4881.

Zhang, J., Dauphas, N., Davis, A.M., Leya, I., Fedkin, A., 2012. The proto-Earth as asignificant source of lunar material. Nat. Geosci. 5, 251–255.