Distribution of p-process 174Hf in early solar system materials and the origin of nucleosynthetic Hf and W isotope anomalies in Ca–Al rich inclusions

Earth and Planetary Science Letters 459 (2017) 70–79

Contents lists available at ScienceDirect

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

Distribution of p-process 174 Hf in early solar system materials and the origin of nucleosynthetic Hf and W isotope anomalies in Ca–Al rich inclusions Stefan T.M. Peters a,b,c,∗ , Carsten Münker a,b , Markus Pfeifer a,b , Bo-Magnus Elfers a,b , Peter Sprung a,b a b c

Institut für Geologie und Mineralogie, Universität zu Köln, Zülpicherstr. 49b, 50674 Cologne, Germany Steinmann-Institut, Poppelsdorfer Schloss, 53115 Bonn, Germany Geowissenschaftliches Zentrum der Georg-August-Universität Göttingen, Department of Isotope Geology, Goldschmidtstrasse 1, 37077 Göttingen, Germany

a r t i c l e

i n f o

Article history: Received 22 July 2016 Received in revised form 1 November 2016 Accepted 4 November 2016 Available online 29 November 2016 Editor: B. Marty Keywords: meteorites supernova injection CAI nucleosynthetic isotope anomalies hafnium tungsten

a b s t r a c t Some nuclides that were produced in supernovae are heterogeneously distributed between different meteoritic materials. In some cases these heterogeneities have been interpreted as the result of interaction between ejecta from a nearby supernova and the nascent solar system. Particularly in the case of the oldest objects that formed in the solar system – Ca–Al rich inclusions (CAIs) – this view is confirm the hypothesis that a nearby supernova event facilitated or even triggered solar system formation. We present Hf isotope data for bulk meteorites, terrestrial materials and CAIs, for the first time including the low-abundance isotope 174 Hf (∼0.16%). This rare isotope was likely produced during explosive O/Ne shell burning in massive stars (i.e., the classical “p-process”), and therefore its abundance potentially provides a sensitive tracer for putative heterogeneities within the solar system that were introduced by supernova ejecta. For CAIs and one LL chondrite, also complementary W isotope data are reported for the same sample cuts. Once corrected for small neutron capture effects, different chondrite groups, eucrites, a silicate inclusion of a IAB iron meteorite, and terrestrial materials display homogeneous Hf isotope compositions including 174 Hf. Hafnium-174 was thus uniformly distributed in the inner solar system when planetesimals formed at the <50 ppm level. This finding is in good agreement with the evidently homogeneous distributions of p-process isotopes 180 W, 184 Os and possibly 190 Pt between different iron meteorite groups. In contrast to bulk meteorite samples, CAIs show variable depletions in p-process 174 Hf with respect to the inner solar system composition, and also variable r-process (or s-process) Hf and W contributions. Based on combined Hf and W isotope compositions, we show that CAIs sampled at least one component in which the proportion of r- and s-process derived Hf and W deviates from that of supernova ejecta. The Hf and W isotope anomalies in CAIs are therefore best explained by selective processing of presolar carrier phases prior to CAI formation, and not by a late injection of supernova materials. Likewise, other isotope anomalies in additional elements in CAIs relative to the bulk solar system may reflect the same process. The isotopic heterogeneities between the first refractory condensates may have been eradicated partially during CAI formation, because W isotope anomalies in CAIs appear to decrease with increasing W concentrations as inferred from time-integrated 182 W/184 W. Importantly, the 176 Lu–176 Hf and 182 Hf–182 W chronometers are not significantly affected by nucleosynthetic heterogeneity of Hf isotopes in bulk meteorites, but may be affected in CAIs. © 2016 Elsevier B.V. All rights reserved.

1. Introduction

*

Corresponding author at: Geowissenschaftliches Zentrum der Georg-AugustUniversität Göttingen, Department of Isotope Geology, Goldschmidtstrasse 1, 37077 Göttingen, Germany. Fax: +49 551 393 982. E-mail address: [email protected] (S.T.M. Peters). http://dx.doi.org/10.1016/j.epsl.2016.11.009 0012-821X/© 2016 Elsevier B.V. All rights reserved.

A long-standing hypothesis for the formation of the solar system is that a shock wave from a nearby supernova event locally compressed the protosolar molecular cloud, causing it to collapse towards the center of its mass (Cameron and Truran, 1977). Such a nearby supernova event may have furthermore provided the solar system with some of its extinct radionuclides

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

(e.g., 26 Al, 41 Ca, 60 Fe; e.g., Sahijpal et al., 1998). However, these particular radionuclides partially may have been produced by irradiation or may have been present in the molecular cloud before it collapsed (Tang and Dauphas, 2012; Gounelle and Meynet, 2012 and references therein). Heterogeneities between meteoritic materials in some stable or long-lived nuclides that were produced in supernovae were interpreted to reflect admixing of ejecta from supernovae into the molecular cloud or into the protoplanetary disk, i.e., in agreement with the supernova trigger hypothesis of solar system formation (e.g., Dauphas et al., 2010; Qin et al., 2011). Particularly anomalous isotope compositions of Ca–Al rich inclusions (CAIs) in primitive meteorites have been interpreted in this respect (McCulloch and Wasserburg, 1978; Brennecka et al., 2013, 2014; Shollenberger et al., 2016), because CAIs are the oldest objects in the solar system and appear to have sampled a largely uniform isotopic reservoir that differed from the bulk solar system composition. However, isotopic heterogeneities between bulk meteorites at least partially reflect the mechanical and/or thermal processing of different presolar nucleosynthetic carrier phases (e.g., Regelous et al., 2008; Trinquier et al., 2009; Mayer et al., 2015), and similar processes may also explain the anomalous compositions of some elements in CAIs (e.g., Akram et al., 2013). Together, there is therefore no unambiguous evidence for the interaction between supernova ejecta and the molecular cloud from which the solar system formed. In order to better understand the origin of nucloesynthetic isotope variations in the early solar system, we analyzed isotope compositions of Hf and W, for the first time involving the pure p-process isotope 174 Hf, in a variety of meteorite types. P-process isotopes are potentially sensitive tracers for nucleosynthetic heterogeneity, because the p-process component of a given element heavier than Fe is typically one to two orders of magnitude less than the s- and r-process components of this element. Importantly, p-process isotopes with A ≥ 168 are predominantly formed in supernovae by photodisintegration reactions during explosive O/Ne shell burning, i.e., the γ -process of nucleosynthesis (e.g., Rauscher et al., 2013 and references therein). Therefore, p-process isotopes with A ≥ 168 are expected to be sensitive monitors for the distribution of supernova material in the early solar system. Recent studies therefore already investigated the distributions of p-process isotopes 180 W (Schulz et al., 2013; Peters et al., 2014; Cook et al., 2014; Holst et al., 2015), 184 Os (Walker, 2012) and 190 Pt (Peters et al., 2015; Hunt and Schönbächler, 2015). No evidence for nucleosynthetic heterogeneity was found in these studies which mainly focused on iron meteorites and thus do not provide a comprehensive view. In this study on 174 Hf, we present the first extensive dataset for a heavy p-process isotope in silicate samples, including CAIs. The latter are of particular interest, because CAIs were previously shown to carry anomalous s- or r-process Hf (Akram et al., 2013). Furthermore, CAIs are known to carry anomalous r- or s-process W (Burkhardt et al., 2008; Kruijer et al., 2014a). To better understand the origin of the Hf isotopic variability between CAIs we therefore also analyzed their W isotope compositions. The distribution of p-process Hf in early solar system materials is of special interest because 176 Hf (∼97% s-process, ∼3% p-process; Klay et al., 1991) is also produced by radioactive decay of long-lived 176 Lu (t 1/2 ∼ 37 Gyr). The 176 Lu–176 Hf decay system is a potentially robust chronometer for dating early solar system processes, but 176 Lu–176 Hf ages for some chondrite, eucrite and angrite samples are too old with respect to the well calibrated Pb–Pb method (e.g., Patchett and Tatsumoto, 1980; Blichert-Toft and Albarède, 1997; Bizzarro et al., 2012; Bast et al., 2017). Heterogeneity of p-process Hf could affect the calculated initial 176 Hf/177 Hf from mineral isochrons and may explain some of the scatter for published ages of different chondrites. Hetero-

71

geneity of Hf isotopes in the early solar system could furthermore affect the 182 Hf–182 W chronometer that is based on the decay of now extinct 182 Hf (t 1/2 ∼ 8.9 Myr). The 182 Hf–182 W decay system has been widely used for the identification of small age differences (ca. >1 Myr for the oldest objects, using recent analytical methods) between metal-silicate segregation processes within early solar system planetary bodies (e.g., Harper and Jacobsen, 1996; Kleine et al., 2004; Schulz et al., 2009; Kruijer et al., 2014b). However, 182 Hf–182 W ages of early solar system objects are typically reported relative to the calculated initial 182 Hf/180 Hf of CAIs (e.g., Burkhardt et al., 2008; Kruijer et al., 2014a). This reasoning is only valid if at time zero the reservoir from which CAIs formed had no significantly different 182 Hf/180 Hf than the remaining solar system. 2. Samples and methods Ordinary H, L and LL chondrites (n = 6, 6, 1), one EL chondrite, two CV chondrites, three eucrite samples and a silicate inclusion that had been separated from the IAB iron meteorite Campo del Cielo (El Taco) (Schulz et al., 2009) were selected for analysis. Mostly equilibrated chondrites (metamorphic types 4–6) were chosen in order to minimize possible bias from residual, undigested presolar grains after digestion. Five terrestrial samples were analyzed including international reference materials BHVO-2 and BCR-2, as well as basalts from the Leinegraben, Germany (BB-46A), and Muriah, Indonesia (Sunda Arc; I12MU, 3MU-13). Additionally, five CAIs were separated from the CV3 chondrites Allende (n = 2; Pfeifer et al., in prep.), NWA 3095 (n = 2; Pfeifer et al., in prep.) and Bali (n = 1; Becker et al., 2015). We also analyzed a CAIdominated sample powder that was previously extracted from Allende by Stracke et al. (2012) (their sample C4). The trace element compositions, mineralogy and petrology of the CAI samples from Allende and NWA 3095 are discussed in detail in Pfeifer et al. (in prep.). For these particular samples, in this manuscript, the original formats of the samples names (CAI_C-BN_#) were shortened to CAI-# to improve the readability. All meteorite samples were cut using a rock saw and crushed in an agate mortar. Typically 0.5–1.0 g of chondrite powder was then digested in a 24 N HF–14 N HNO3 –9.5 N HClO4 mixture (∼5:5:2) at 180 ◦ C on a hotplate. Terrestrial basalts and eucrites (typically 0.1–0.3 g) were digested in a ca. 1:1 mixture of 24 N HF–14 N HNO3 . The silicate inclusion of Campo del Cielo contained high amounts of dark material that remained after table top digestion, possibly graphite, and the sample was therefore digested in Parr autoclaves in 1:1 24 N HF–14 N HNO3 at 180 ◦ C for three days, after which virtually all material was digested. The CAI samples were carefully drilled out of the meteorite matrices while avoiding direct drilling into the CAI. Any matrix material that had been drilled out was then separated from the samples by handpicking under a binocular microscope. The CAI samples were digested in a ∼3:1 mixture of 24 N HF–14 N HNO3 mixture at 120 ◦ C, after which residual refractory mineral grains were attacked with perchloric acid at 180 ◦ C. For all samples except CAIs and the LL chondrite Khanpur, Hf was directly separated from the rock matrix using EichromTM Ln resin (Münker et al., 2001). The Hf cuts were subsequently further purified from Yb using BioRad AG1-X8 anion exchange resin in order to minimize interfering 174 Yb (Peters et al., 2015). The CAIs and LL chondrite Khanpur were processed over Ag50W × 8 (200–400 mesh) cation exchange resin in order to also separate W for isotope analysis. For these samples the HFSE, W, and Ti elute from the resin immediately upon loading in 1 M HCl–0.1 M HF (Patchett and Tatsumoto, 1981). After having passed through an Ag1 × 8 (200–400 mesh) anion column that removed remaining matrix elements and Ta, W was subsequently separated from Hf, Ti, Zr using columns using EichromTM TEVA resin (Peters et al., 2015).

72

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

Hafnium was subsequently purified from Ti and Zr on EichromTM Ln resin following Bast et al. (2015), but using 6 M HCl–0.06 M HF to separate Zr and Hf. Hafnium was then eluted in 2 M HF (Münker et al., 2001). All samples were analyzed using the Thermo FinniganTM Neptune MC-ICP-MS at the joint Cologne–Bonn facility. For Hf isotope analysis, the following isotopes were collected: 172 Yb (monitor for interfering 174 Yb, 176 Yb), 174 Hf, 175 Lu (monitor for isobaric 176 Lu), 176 Hf, 177 Hf, 178 Hf, 179 Hf, 180 Hf, 182 W (monitor for isobaric 180 W). Two Faraday amplifiers with 1012  feedback resistors were used for the collection of the 174 Hf and 172 Yb isotope beams. Using these amplifiers and ion currents above >5 × 10−13 A on 174 Hf, measurement precisions for interference and mass bias corrected 174 Hf/177 Hf ratios were typically <70 ppm for matrixfree solutions. The minor isobaric interference by the rare isotope 180 Ta (180 Ta/181 Ta ∼10−4 ) on 180 Hf was not monitored online but checked for in a ∼5% aliquot that had been diluted by a factor ∼5 prior to analysis. Interfering 180 Ta was always sufficiently low (<10−13 A on 181 Ta in the analyte). Regular sampler cones and x-skimmer cones were used throughout: sampler cones with a larger aperture (“Jet” cones) were also tested, but measurements in this setup yielded spurious isotope ratios that were apparently dependent on Zr/Hf (Peters et al., 2015). After ion exchange chromatography, Yb/Hf (element ratio) was typically <10−5 . Tests with Yb-doped Hf solutions showed that at this level, isobaric 174 Yb can be accurately corrected for (Peters et al., 2015). Moreover, potential cosmogenic, radiogenic, and/or nucleosynthetic isotope variability in Yb is expected to have negligible effect on the interference corrected Hf isotope ratios at this low level of Yb/Hf. For W isotope analysis the following isotopes were collected: 178 Hf (monitor for isobaric 180 Hf), 180 W, 181 Ta (monitor for isobaric 180 Ta), 182 W, 183 W, 184 W, 186 W and 188 Os (monitor for isobaric 184 Os, 186 Os). The low-abundance isotope 180 W and interference monitor 178 Hf were collected using the Faraday cups with 1012  feedback resistors (Schulz et al., 2013; Peters et al., 2014, 2015). Although W isotope data were robust for the major isotopes, we suspect that 180 Hf-interference corrected 180 W/i W ratios were biased due to an unidentified interfering species on 178 Hf, because the interference corrected 180 W/i W isotope ratios were up to several hundreds of ppm lower than the range that was previously observed in natural samples. Importantly, during measurements in a companion study on the Ta isotope compositions of these samples, the same sample solutions showed unusually low 180 Hfinterference corrected 180 Ta/181 Ta by several hundreds of ppm. Also in this case 178 Hf was the monitoring isotope (Pfeifer et al., accepted). A possible reason for the suggested interference on mass 178 may be organics that were released from the cation exchange resin during ion exchange chromatography, an issue that did not affect previous studies because these used a different separation scheme that did not include cation exchange resin (Schulz et al., 2013; Peters et al., 2014). All Hf data are reported as the deviation from the composition of AMES Hf-metal in parts per million using 177 Hf in the denominator, i.e.,



μi Hf = 106

(i Hf/177 Hf)Sample −1 (i Hf/177 Hf)AMES



and using the exponential law for correction of the instrumental mass bias normalizing to 177 Hf/179 Hf = 0.7325. Tungsten isotope data are reported in a similar notation, i.e., as μ183 W, the deviation of 183 W/184 W relative to the NIST SRM 3163 W-solution (isotopically indistinguishable from AMES W-metal; Peters et al., 2014) in ppm, using 186 W/184 W = 0.92767 and the exponential law for correction of the instrumental mass bias. Measurement uncertainties are given as 2 S.D. of the standard measurements throughout an analytical session (n > 10 standards per analytical session) or

as the within-run variation of the sample (2 S.E.M.) if the latter is larger. 3. Results Measured Hf isotope compositions are listed in Table 1 and displayed in Fig. 1. For individual samples of terrestrial basalts μ174 Hf, μ178 Hf, and μ180 Hf are indistinguishable from the AMES Hf-metal solution, although the population means are slightly lower (μ174 Hf = −24 ± 13 ppm) and higher (μ178 Hf = +7 ± 3; μ180 Hf = +8 ± 7; all 95% confidence intervals) than the AMES Hfmetal solution, respectively. For μ178 Hf these data are in excellent agreement with the data by Sprung et al. (2010) who also obtained a slightly positive average μ178 Hf of +7 ± 3 ppm for terrestrial samples relative to AMES Hf-metal. For μ180 Hf our data overlap with those by Sprung et al. (2010) (μ180 Hf = −3 ± 7 ppm) within the level of uncertainty. Sprung et al. (2010) did not report μ174 Hf values. Most chondrite samples show indistinguishable μ174 Hf, μ178 Hf and μ180 Hf from the AMES Hf-metal solution. However, the EL chondrite Pillistfer displays slightly elevated μ174 Hf and μ178 Hf of +130 ± 89 ppm and +15 ± 4 ppm, but 36 ± 6 ppm lower μ180 Hf values than the AMES Hf-metal solution. Elevated μ178 Hf and negative μ180 Hf values of similar magnitude are also observed for Sierra Colorada and Wagon Mound, L5 and L6 chondrites, respectively. For Wagon Mound these data were reproduced for a digestion replicate of the same sample powder. The average Hf isotope composition of all other chondrites is indistinguishable from the terrestrial samples within uncertainty (μ174 Hf = +5 ± 29; μ178 Hf = +6 ± 4; μ180 Hf = −1 ± 5). Eucrite samples have somewhat variable Hf isotope compositions, but are on average indistinguishable from chondrites, terrestrial samples and AMES Hfmetal (μ174 Hf = +36 ± 57; μ178 Hf = +3 ± 4; μ180 Hf = −3 ± 4). Similar to Pillistfer, the silicate inclusion of Campo del Cielo IAB iron meteorite displays an elevated μ174 Hf of +128 ± 92 ppm as well as positive μ178 Hf and negative μ180 Hf of +16 ± 5 and −36 ± 11 ppm, respectively. In contrast, all CAIs display negative μ174 Hf between −428 ± 22 ppm and −102 ± 88 ppm, except for CAI-1, for which the measurement uncertainty is too large to yield a decisive result. Except for CAI-1 and CAI-4, all CAIs show slightly positive μ180 Hf of <40 ppm, and all except CAI-1 and Bali-CAI show positive μ178 Hf of ≤20 ppm. CAI-4 shows a tentatively negative μ180 Hf of −3 ± 1 ppm as the weighted mean of two analyses. The positive μ180 Hf and μ178 Hf observed in most CAIs are in good agreement with the data by Akram et al. (2013), in which μ174 Hf was not reported. Furthermore, Akram et al. (2013) also reported negative μ180 Hf in one CAI, i.e. similar to what is observed in this study for CAI-4. The LL chondrite Khanpur has indistinguishable μ183 W from the NIST SRM 3163 W-solution, The CAIs show variably positive μ183 W of ca. 10–35 ppm except for CAI-4, which has indistinguishable μ183 W from the NIST SRM 3163 W-solution. The range of μ183 W in these CAIs is similar to that previously reported for coarse-grained CAIs (Burkhardt et al., 2008; Kruijer et al., 2014a). 4. Discussion 4.1. Effects of cosmic ray exposure The measured Hf isotope compositions of the three anomalous chondrite samples agree reasonably well with the model for secondary neutron capture effects by Sprung et al. (2010), i.e., the negative μ180 Hf of these samples are coupled to elevated μ178 Hf and, in the case of Pillistfer, also elevated μ174 Hf (Fig. 2A, B). This is also the case for the silicate inclusion of Campo del Cielo, including the elevated value for μ174 Hf. Secondary neutron capture reactions may occur due to sample exposure to cosmic rays.

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

73

Table 1 Hf and W isotope compositions of terrestrial and extraterrestrial silicate samplesa .

μ174 Hf

μ178 Hf

μ180 Hf

−7 ± 79 40 ± 100 80 ± 92 −12 ± 78

19 ± 9 15 ± 11 4±9 6±7

21 ± 22 9 ± 22 6 ± 18 16 ± 19

19 ± 43 −29 ± 54 −6 ± 56 −68 ± 92 −28 ± 96 −50 ± 100 −21 ± 39 −61 ± 88 −30 ± 78

10 ± 4 11 ± 12 4±8 0±9 2 ± 13 9 ± 11 3±5 −1 ± 6 10 ± 12

12 ± 10 4 ± 19 12 ± 22 −5 ± 22 18 ± 23 9 ± 22 8 ± 11 −9 ± 11 27 ± 19

65 ± 56 53 ± 94 60 ± 100

4±5 9 ± 13 7±5

−10 ± 14 −15 ± 22 1 ± 19

28 ± 69 −60 ± 110 90 ± 140 54 ± 74 −40 ± 100 40 ± 100 18 ± 56 0 ± 200 −21 ± 99 −100 ± 140 102 ± 74 79 ± 67

2±5 −5 ± 15 4±6 12 ± 4 2 ± 11 24 ± 4 14 ± 12 5 ± 19 7±6 5±5 26 ± 6 22 ± 12

−6 ± 11 −1 ± 19 0 ± 15 −3 ± 19 2 ± 14 −24 ± 13 2 ± 22 14 ± 19 6±9 7 ± 12 −43 ± 14 −40 ± 9

46 ± 50 7 ± 52 130 ± 89 −50 ± 140 2 ± 56

21 ± 5 2±5 15 ± 4 1±5 −3 ± 4

−41 ± 8 −6 ± 6 −36 ± 6 0±9 13 ± 15

−11 ± 57 70 ± 100 130 ± 100 60 ± 140

2±3 6±4 7±4 7±5

0±9 −11 ± 11 −3 ± 5 4±5

Weighted mean Stannern

90 ± 66 −31 ± 93 8 ± 56 53 ± 41

6±3 0±4 10 ± 9 10 ± 4

−1 ± 3 −16 ± 9 0±9 −3 ± 9

Weighted mean

30 ± 31

6±3

−7 ± 5

Terrestrial samples BHVO-2a

Basalt

Weighted mean BCR-2b BB-46Ab

Basalt Basalt

I12MU 3MU-13b

Basalt Basalt

Chondrites Tulia(a) Pulltuskb

H3–4 H5

Weighted mean NWA 926 Yelland Dry Lake NWA 4296 NWA 4009 Sierra Colorada NWA 904 NWA 515b Gilgoin NWA 4292 Wagon Mound

H4 H4 H4 H5–6 L5 L5 L6 L6 L6 L6

Weighted mean Khanpur Pillistfer Vigarano NWA 3118b

LL5 EL6 CV3 CV3

Eucrites Millbillillie NWA 4051

Iron meteorites Campo del Cielo (El Taco)

IAB (inclusion)

130 ± 90

16 ± 5

−36 ± 11

Ca–Al rich inclusions CAI_C-BN_1 CAI_C-BN_3 CAI_C-BN_4

Allende Allende NWA 3095

200 ± 990 −300 ± 360 −406 ± 82 −449 ± 76

0 ± 20 20 ± 16 1±8 4±5

13 ± 35 30 ± 21 −2 ± 6 −4 ± 5

Weighted mean CAI_C-BN_5 CAI_C-BN_6c Bali-CAId

NWA 3095 Allende Bali

−428 ± 22 −151 ± 71 −102 ± 88 −200 ± 78

3±2 17 ± 5 17 ± 5 3±8

−3 ± 1 36 ± 7 25 ± 8 32 ± 9

μ183 W

μ182 W

10 ± 11

−214 ± 8

28 ± 16 25 ± 13 6±8

−84 ± 36 −60 ± 19 −157 ± 10

−7 ± 10 11 ± 16 −20 ± 20

−193 ± 13 −138 ± 24 −30 ± 25

a Given as deviations in ppm of the i Hf/177 Hf and i W/184 W ratios from the AMES Hf-metal solution and NIST SRM 3163 W-solution, respectively, using 0.7325 and 186 W/184 W = 0.92767 for normalization of the instrumental mass bias.

177

Hf/179 Hf =

b

Some data previously published in Peters et al. (2015). Sample C4 from Stracke et al. (2012). d W isotope data from Becker et al. (2015). Measurement uncertainties are given either as 2 S.D. of the standard measurements throughout an analytical session (n > 10 standards per analytical session), or as the within-run variation of the sample (2 S.E.M.), whichever is larger. c

For Hf isotopes, secondary neutron capture effects were previously not resolved in chondrites for isotope ratios involving 177 Hf, 178 Hf, 179 Hf and 180 Hf, but they were recognized in mesosiderites (Sprung et al., 2010) and lunar samples (Sprung et al., 2013; Gaffney and Borg, 2014. The most probable neutron capture re-

actions affecting Hf isotope compositions occur at epithermal neutron energies and are 177 Hf(n, γ )178 Hf and 178 Hf(n, γ )179 Hf with resonance integrals of ∼7173 barn and 1910 barn, respectively (Mughabghab, 2003). The anomalous Hf isotope composition of the silicate inclusion from Campo del Cielo may therefore re-

74

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

Fig. 1. Non-radiogenic Hf isotope ratios in terrestrial and extraterrestrial silicate samples given as μi Hf, i.e., the deviation of i Hf/177 Hf from the AMES Hf-metal solution (black line) in ppm, using 179 Hf/177 Hf = 0.7325 and the exponential law for correction of the instrumental mass bias. Shaded bars correspond to 95% confidence intervals of the population means. For calculation of the 95% confidence limits for chondrites data were excluded for Wagon Mound (L), Sierra Colorada (L) and Pillistfer (EL), because these samples were probably affected by secondary neutron capture reactions (see text).

late to the high Fe-content of this meteorite, because neutron energy spectra in Fe-rich target materials are modified towards higher proportions of epithermal relative to thermal neutrons (Kollar et al., 2006). Our observation is in line with what was observed for mesosiderites (Sprung et al., 2010) using Hf and Sm isotope data. If the samples that were probably affected by neutron capture reactions are excluded from the dataset, all extraterrestrial samples except for CAIs have Hf isotope compositions that do not differ from the group average of terrestrial samples. 4.2. Distribution of p-process 174 Hf on an asteroidal scale Bulk chondrites from different groups that lack secondary neutron capture effects, and eucrites have indistinguishable μ174 Hf from terrestrial samples. The grand weighted mean for μ174 Hf from different sample groups is −8 ± 23 ppm and defines the μ174 Hf of the inner solar system. We can furthermore confirm the conclusion by Sprung et al. (2010) that s- and r- process Hf are uniformly distributed in the inner the early solar system at the ppm level, and have now complemented their sample suite with data for eucrites. The uniform distribution of 174 Hf in bulk meteorites is consistent with the homogeneity of pprocess 180 W, 184 Os and possibly 190 Pt in different iron meteorite groups at levels of <100 ppm, <2700 ppm, and <400 ppm, respectively (Walker, 2012; Peters et al., 2014, 2015; Hunt and Schönbächler, 2015). Altogether, these data indicate that if a supernova had injected p-nuclides into the nascent solar system, these nuclides must have become efficiently homogenized by the time at which planetesimals formed. This conclusion is in apparent agreement with evidence from Ni isotopes, arguing that extant supernova-derived 60 Fe (t 1/2 ∼ 2.6 Myr) was homogeneously distributed at the level of <10% by the time when planetesimals formed (Dauphas et al., 2008; Tang and Dauphas, 2012). However, it is still unclear as to whether minor variations in 60 Fe/56 Fe may have existed (e.g. Mishra and Chaussidon, 2014).

4.3. Nucleosynthetic isotope anomalies in CAIs 4.3.1. Nucleosynthetic isotope heterogeneity between CAIs In all CAIs except CAI-4, elevated μ178 Hf and μ180 Hf indicate the presence of r-process depleted (or s-process enriched) Hf, in agreement with the findings of Akram et al. (2013) (Fig. 2C). In contrast, CAI-4 carries comparatively r-process enriched (s-process deficient) or terrestrial Hf. The contrasting compositions of these CAIs suggest the presence of two isotopically distinct components in the CAI forming region with respect to Hf, in agreement with Akram et al. (2013) who suggested the existence of two or possibly three such components. The negative μ174 Hf in all CAIs including CAI-4 can only be explained by a deficit in p-process Hf, i.e., an isolated depletion in 174 Hf (Fig. 2D). Therefore, a complementary, third component that was enriched in p-process Hf is needed to explain the average Hf isotope composition of the inner solar system. If μ174 Hf in CAIs is corrected for r-process heterogeneity using the measured μ180 Hf and the s-process yields from Wisshak et al. (2006), the corrected μ174 Hfcorr for CAI-4 does still not overlap with three other CAIs (Table 2). The correction becomes somewhat more significant when the s-process yields from Arlandini et al. (1999) are considered rather than those by Wisshak et al. (2006), but still yields a different μ174 Hfcorr for CAI-4 than for the other CAIs (Table 2). Importantly, the s-process yields by Wisshak et al. (2006) are more consistent with the coupled variations between μ178 Hf and μ180 Hf in these samples (Fig. 2C) and are therefore preferred as the more realistic of the two sets. The corrections for s-process heterogeneity are negligible for both models on the scale of the observed μ180 Hf. Consequently, the CAI forming region was probably heterogeneous with respect to p-process Hf in addition to r- (or s-)process Hf. The p-process enriched component in the solar system therefore may have been sampled by CAIs, but at lower proportions than by chondrites. In fact, μ174 Hfcorr and μ180 Hf in our CAI samples exhibit a tentative co-variation (Fig. 3). The three components that potentially contributed Hf to the inner solar system are therefore relative to its bulk composition: (I) depleted in rprocess Hf (alternatively: enriched in s-process Hf) and depleted in

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

75

Fig. 2. (A, B) μ180 Hf vs. μ178 Hf and μ174 Hf for bulk chondrites (green), eucrites (red) and the silicate inclusion of Campo del Cielo (orange). Symbols are identical to Fig. 1. Most chondrite and eucrite samples scatter around the composition of the terrestrial standard, but Sierra Colorada (SC), Wagon Mound (WM), Pillistfer (EL) and also the silicate inclusion of Campo del Cielo (IAB) have elevated μ174 Hf and μ178 Hf, but lower μ180 Hf than the other samples. These coupled isotope anomalies are in good agreement with the model for secondary neutron capture effects by Sprung et al. (2010) that is depicted by the blue arrows (upper and lower directions correspond to neutron energy spectra with epithermal-to-thermal ratios of 1 and 8, respectively). (C) Coupled μ178 Hf and μ180 Hf anomalies in CAIs from Allende (circles), NWA 3095 (diamonds) and Bali (square) cannot be explained by neutron capture effects, but suggest a different r-, s-process Hf composition in CAIs than the terrestrial standard. Note that μ178 Hf and μ180 Hf between CAI-4 (indicated) and the other CAIs is different, suggesting heterogeneity of either r-, or s-process Hf between CAIs, in agreement with the data by Akram et al. (2013) (depicted in grey). (D) Negative μ174 Hf in CAIs can only be explained by a deficit in p-process Hf, but neither r- nor s-process heterogeneity (light and dark ranges, respectively) can fully explain the variation in μ174 Hf between CAI-4 (indicated) and three of the other CAIs, suggesting heterogeneity of p-process Hf between different CAIs. The width of the light and dark ranges corresponds to the measurement uncertainty of μ174 Hf in CAI-4. Models for nucleosynthetic effects are based on the s-process yields from Wisshak et al. (2006) and are compared in Fig. 2C with the expected nucloesynthetic effects based on the s-process yields by Arlandini et al. (1999). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 2 μ174 Hf and

μ182 W in Allende and Bali CAIs corrected for r-process heterogeneitya .

Model CAI_C-BN_1 CAI_C-BN_3 CAI_C-BN_4 CAI_C-BN_5 CAI_C-BN_6 Bali-CAI

μ174 Hfcorr

μ174 Hfcorr

μ182 Wcorr

Wisshak et al. (2006)

Arlandini et al. (1999)

Burkhardt and Schönbächler (2015)

160 ± 990 −380 ± 360 −419 ± 22 −257 ± 71 −176 ± 88 −295 ± 78

150 ± 990 −410 ± 360 −416 ± 22 −285 ± 71 −195 ± 88 −319 ± 78

−64 ± 36 −42 ± 19 −153 ± 10 −198 ± 13 −130 ± 24 −44 ± 25

a Based on the measured μ180 Hf and μ183 W using s-process yields (Hf; Wisshak et al., 2006; Arlandini et al., 1999) and the isotope compositions of leachates from primitive meteorites (W; Burkhardt and Schönbächler, 2015).

p-process Hf, (II) enriched in r-process Hf (alternatively: depleted in s-process Hf) and depleted in p-process Hf, and (III) enriched in p-process Hf with unknown r–s-process proportions (Fig. 3). Based on their nucleosynthetic compositions, we suggest that component

I carried nucleosynthetic anomalies resulting from AGB star material, whereas components II and III carried an overabundance of supernova materials compared to the bulk composition of the solar system.

76

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

Fig. 3. μ174 Hf in CAIs corrected for r-process heterogeneity (μ174 Hfcorr ) based on Wisshak et al. (2006) against μ180 Hf in comparison to the average composition of the inner solar system (this study, Sprung et al., 2010). At least two isotopically distinct components are required to explain the variability in μ180 Hf between CAIs. A third component is required to explain the different μ174 Hfcorr of CAIs and the average composition of the inner solar system. According to the apparently variable μ174 Hfcorr and a tentative co-variation between μ174 Hfcorr and μ180 Hf, this third component may also have been sampled by CAIs. Compared to the average composition of the inner solar system, these three components with distinct Hf isotope compositions are (I) r-process depleted (s-process enriched) and p-process depleted; (II) r-process enriched (s-process depleted) and p-process depleted; (III) p-process enriched with unknown r- and s-process composition.

The positive μ183 W in all CAIs except CAI-4 and the Bali-CAI suggest that CAIs predominantly carried r-process enriched (sprocess depleted) W, i.e., opposite to what is observed for Hf, in agreement with Burkhardt et al. (2008) and Kruijer et al. (2014a). In all samples except CAI-4, μ183 W and μ180 Hf appear to be negatively correlated, suggesting that components I and II with distinct Hf isotope compositions also had different W isotope compositions (Fig. 4). CAI-4 seems to have tapped a third component that had a W isotope composition close to that of the bulk solar system with respect to r- and s-process W, similar to what is observed for the Hf isotope composition of this sample. Together, three components with distinct combined Hf and W isotope compositions are therefore required to explain the isotopic inventory of the CAIs. This feature is difficult to explain by heterogeneously distributed supernova ejecta, which rather would result in only two isotopically distinct components. In fact, the composition of component III relative to components I and II differs markedly from the modeled effects for r- or s-process heterogeneity (Fig. 4). The combined Hf and W isotope compositions of CAIs may therefore rather reflect different nebular processing of r-process Hf versus r-process W carrier phases prior to CAI formation. 4.3.2. Nucleosynthetic isotope anomalies in CAIs and the supernova injection scenario As outlined above, the combined Hf and W isotope compositions of CAIs require that selective processing of presolar carrier phases occurred prior to or during CAI formation. It is plausible that such processes led to isotopic heterogeneities in the solar system beyond the scale of the CAI forming region and also caused the isotope anomalies in other elements in CAIs, e.g., in Sr, Mo, Ru, Ba, Nd, Sm, Gd, Dy, Er and Yb (Brennecka et al., 2013, 2014; Shollenberger et al., 2016). Instead of such a model, the latter studies suggested that the isotope compositions of these elements are best explained by an injection of supernova nucleosynthetic material into a reservoir that was isolated from the CAI forming region in space or time, because the isotope compositions of these ele-

Fig. 4. Coupled μ180 Hf and μ183 W data for most CAIs show an apparent trend of increasing μ180 Hf with decreasing μ183 W, except for CAI-4 that has similar μ180 Hf and μ183 W to the average composition of the inner solar system. The coupled μ180 Hf and μ183 W cannot be explained by heterogeneity of r-, s-process Hf relative to the average composition of the inner solar system (indicated), and requires a component with markedly different r-process Hf/r-process W ratio than the relative production factors for these elements during stellar nucleosynthesis. The value for μ180 Hf of the inner Solar System is defined by this study and by Sprung et al. (2010), whereas μ183 W of the inner Solar System is defined by Budde et al. (2016) and is confirmed by our data for the LL chondrite Khanpur. Note that the dashed, blue lines are drawn to demonstrate the need of three isotopically distinct components in the CAI forming region, and do not necessarily imply mixing relations, which rather would be curves instead. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

ments were uniform in the CAIs that they studied. Conversely, our data suggest that Hf and W isotopes in CAIs are variable, in agreement with Akram et al. (2013) (Hf, Zr), Kruijer et al. (2014a) (W), Leya et al. (2009) (Ti) and Quitté et al. (2007) (Ni), and possibly also with the Mo isotope data by Burkhardt et al. (2011). Moreover, recent data from Bouvier and Boyet (2016) suggest that Sm and Nd isotopes in CAIs are possibly variable as well. It is therefore conceivable that CAIs formed in an isotopically heterogeneous nebula with respect to Hf, W, Zr and Ti, and possibly also Sm, Nd and Mo. We propose a three-stage model for the origin of isotope anomalies of these elements in CAIs: (I) Prior to formation of the first condensates, thermal and/or mechanical processing of presolar nucleosynthetic carrier phases had left behind an isotopically heterogeneous nebula. As a result, the CAI-forming region was on average enriched in different carrier phases with different isotope compositions than the bulk solar system, and was likely isotopically heterogeneous. (II) In the CAI forming region, refractory condensates formed during multiple evaporation and subsequent condensation events. The latter suggestion is in good agreement with petrologic and trace element evidence for the formation of multiple generations of refractory condensates (e.g., Ireland and Fegley, 2000; Ivanova et al., 2012; Krot et al., 2014 and references therein) and is needed to explain the isotopic variability between CAIs. Such evaporation events affecting nucloesynthetic carrier phases may have occurred either contemporaneously in localized parts of the nebula, but may also have happened at multiple times, e.g., due to the admixing of presolar dust in the CAI forming region or by expansion of its size. (III) Coarse-grained CAIs formed by coagulation and fragmentation processes of the earliest condensates (Charnoz et al., 2015), by which their isotope compositions were partially homogenized. The latter process is supported by a correlation of μ183 W with μ182 W for the Allende CAIs, i.e., a proxy for their Hf/W and possibly W concentrations (Fig. 5A, B; Table 2 displays μ182 W cor-

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

77

Fig. 5. (A) μ183 W in CAIs as a proxy for nucleosynthetic heterogeneity versus measured (open symbols) and r-, s-process heterogeneity corrected (closed symbols) μ182 W as a proxy for their Hf/W. For the CAIs from Allende and NWA 3095 (circles and diamonds, respectively), μ183 W correlates with μ182 Wcorr , suggesting that the highest nucleosynthetic isotope anomalies occur in CAIs with the most radiogenic 182 W, thus in those CAIs that have the highest Hf/W. The CAI from Bali (square) possibly plots below the trend for Allende CAIs due to different initial Hf/W values. (B) No significant correlation is seen between μ180 Hf and μ182 Wcorr , suggesting that variations in μ183 W are predominantly dependent on W concentrations. Corrections for nucleosynthetic heterogeneity are based on the s-process yields by Wisshak et al. (2006).

rected for nucleosynthetic heterogeneity). Such an isotopic homogenization process could explain why the W isotope compositions of fine-grained CAIs that probably formed by condensation define a larger compositional range than those of coarse-grained CAIs (Kruijer et al., 2014a), because fine-grained CAIs were less subject to these homogenization processes. Brennecka et al. (2013) pointed out that light r-process ( A ≤ 130) Sr, Mo, Zr, and Ba nuclides in CAIs are decreasingly enriched with increasing mass, whereas heavy r-process ( A > 130) Sm and Nd nuclides in CAIs are increasingly depleted with increasing mass. They ascribed this apparent trend to a supernova injection of heavy r-process nuclides into a reservoir that remained untapped by CAIs. In the context of our model, carrier phases of light r-process Sr, Zr, Mo, Ru, Ba, i.e., elements that show r-process excesses in CAIs, had become enriched in the CAI forming region by thermal and/or mechanical processing before CAI formation. In contrast, carrier phases of heavy r-process Sm, Nd, Gd, Dy, Er, and Yb, i.e., elements that predominantly show r-process deficits in CAIs, had become predominantly depleted. The collective depletion of heavy r-process Sm, Nd, Gd, Dy, Er, and Yb in CAIs may therefore rather reflect the chemical similarities between these elements, which are all lanthanides that probably resided in presolar carrier phases with similar physicochemical properties. Likewise, light r-process Zr and Mo nuclides that are predominantly enriched in CAIs were probably also carried to large extent by similar presolar phases (Zr and Mo are members of the extended high field strength element group; HFSE). Also light r-process Sr and Ba nuclides that are supposedly enriched in CAIs (e.g., Brennecka et al., 2013) were possibly carried by presolar phases with similar physicochemical properties, because Ba and Sr are chemically similar elements (both are alkaline earth metals). Although our model can therefore explain the similarity in r-process enrichment or depletion for some elements, it does not explain the apparent decrease in r-process excess and depletion with increasing mass for the light and heavy r-process isotopes, respectively (Brennecka et al., 2013). However, if true, this trend would also result in resolvable Te isotope anomalies ( A ∼ 128) in CAIs that are either not present or rather reflect an r-process deficit (s-process excess) (Fehr et al. 2009; Moynier et al., 2009). Moreover, the isotope compositions of Gd, Dy, Er, Yb and W in CAIs do not well fit the trend of increasing deple-

tion with mass for the heavy r-process nuclides, while Hf shows both depletions and enrichments (this study, Akram et al., 2013). We therefore consider the selective processing of presolar carrier phases a robust alternative to the supernova injection scenario by Brennecka et al. (2013) to explain the isotopic differences between CAIs and the average inner solar system composition. According to our model, the isotope composition of most elements in chondrites (e.g., Hf, W) does likely not represent a distinct reservoir that is complementary to the CAI forming region (cf. Brennecka et al., 2013), but rather reflects multiple blends of different nucleosynthetic carrier phases that had become fully homogenized by the time of chondrite formation. We note that local enrichments of the least abundant W carrier phases before chondrite formation may also have been preserved in other early objects than CAIs, e.g., some iron meteorite groups show W isotope anomalies similar to CAIs (i.e., the IVB and IID groups and the ungrouped irons Chinga and Deep Springs; e.g., Qin et al., 2008; Kruijer et al., 2012, 2014b), and so do CV chondrules (Becker et al., 2015; Budde et al., 2016). We also note that the isotopic carrier phases of some other elements, e.g., Ti and Mo were evidently heterogeneously distributed when the chondrite parent bodies formed (e.g., Trinquier et al., 2009; Burkhardt et al., 2011). If the solar system became increasingly homogenized with time, it is therefore expected that CAIs sampled a nebula that was even more isotopically heterogeneous with respect to these elements. Indeed, the range of Ti isotope compositions in CAIs is ca. two times larger than the range observed for chondrites, and some CAIs even yield the most anomalous Ti isotope compositions of any object from solar system origin (Trinquier et al., 2009; Leya et al., 2009). For Mo, the data by Brennecka et al. (2013) suggest a uniform excess of r-process nuclides in the CAI forming region, but Burkhardt et al. (2011) reported one CAI with a large deficit in s-process Mo, similar but a factor >4 greater than the s-process deficits that are found in the most anomalous bulk meteorites. In our model, the nucloesynthetic carrier phases with r-process excess Mo that were tapped by the majority of CAIs contributed only a minor fraction of Mo to the inner solar system, whereas s-process depleted and s-process enriched carrier phases (Burkhardt et al., 2012) contributed a more significant fraction. Consequently, on the scale of the inner solar system, the

78

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

r-process excess was rapidly eradicated by homogenization processes, whereas s-process depleted and s-process enriched carrier phases had not yet fully been homogenized by the time when the chondrite parent bodies formed. Data for more CAIs, preferably for fine-grained CAIs, are needed to verify whether the Mo isotope carrier phases were indeed heterogeneously distributed in the solar system at the time of CAI formation. 4.4. Implications for 176 Lu–176 Hf and 182 Hf–182 W chronology The now established uniform distribution of p-process 174 Hf on larger inner solar system bodies implies that also the suggested 3% p-process contribution to 176 Hf was homogeneous at the level of <50 ppm. Consequently, 176 Hf/177 Hf between different parent bodies may have been affected by nucleosynthetic heterogeneity at <1.5 ppm, negligible with respect to the uncertainty of the calculated initial 176 Hf/177 Hf from isochron relationships. The p- and r-, (or s-) process heterogeneity in CAIs, in contrast, may influence 176 Lu–176 Hf systematics and must hence be taken into account. For example, depletion of 100 ppm r-process Hf increases the measured (i.e., mass bias corrected) 176 Hf/177 Hf by 92 ppm (0.92ε -units), whereas a depletion of 400 ppm p-process Hf decreases 176 Hf/177 Hf by 12 ppm, using 179 Hf/177 Hf = 0.7325 for normalization. For the extreme and unlikely scenarios in which 182 Hf was exclusively produced either by the r- or s-process (see e.g., Lugaro et al., 2014), 100 ppm variability in r-, s- process Hf corresponds to a different 182 Hf/180 Hf of 86 ppm, i.e., well within the level of uncertainty of the most precise initial 182 Hf/180 Hf of CAIs reported by Kruijer et al. (2014a) (182 Hf/180 Hf = 1.018 ± 0.043 × 104 ). Furthermore, for objects with an age of 4.567 Ga, a different 182 Hf/180 Hf of 86 ppm would result in a measured difference for 182 W/184 W of <0.1 ppm at four times the chondritic 180 Hf/184 W ratio, i.e., corresponding to the highest Hf/W ever observed in CAIs (Kruijer et al., 2014a). Hafnium isotope heterogeneities of the magnitudes that were observed in this study therefore have negligible effect on the inferred initial 182 Hf/180 Hf and 182 W/184 W of CAIs, i.e., 182 Hf–182 W ages of early solar system objects are unaffected at the present levels of analytical uncertainty. It is, however, possible that the finegrained CAIs that were used to constrain the initial 182 Hf/180 Hf carry more anomalous Hf than the CAIs that were studied here, because they also contained more anomalous W, by up to a factor ∼15–20 (Kruijer et al., 2014a). 5. Conclusions The abundances of heavy p-process isotopes ( A > 130) are potentially sensitive tracers for putative heterogeneity of supernova ejecta between different early solar system materials. Between Earth, different chondrite groups and eucrites p-process 174 Hf is uniformly distributed at <50 ppm, in good agreement with the homogeneous distributions of p-process 180 W, 184 Os and possibly 190 Pt between iron meteorite groups. If a supernova event would have injected p-nuclides into the young solar system, these nuclides therefore must have become largely homogenized by the time at which planetesimals formed. In contrast to bulk meteorites, Ca–Al rich inclusions formed from a heterogeneous reservoir with respect to p- and r- (or, alternatively, s-) process Hf, and also with respect to r- (or, alternatively, s-) process W. Their combined Hf and W isotope compositions suggest that these CAIs sampled at least three isotopically different components. In at least one of these components, the proportion of r- and s-process derived Hf and W deviates from that of supernova ejecta. The variable Hf and W isotope compositions of CAIs are therefore best explained by the selective processing of presolar nucleosynthetic carrier phases prior to CAI formation. Likewise, the isotope anomalies found for

other elements in CAIs relative to the average isotope composition of the solar system may have resulted from nebular processing of presolar carrier phases, and not from a supernova injection that was isolated from the CAI forming region in space or time. Variable nucloesynthetic isotope compositions between CAIs may have originated from localized evaporation and condensation events or, alternatively, from the admixing of nucloesynthetic carrier phases to the CAI forming region during the period of CAI formation. Homogenization of the CAI forming region subsequently occurred as CAIs grew in mass by e.g., coagulation and fragmentation processes, as is suggested by an apparent decrease in anomalous W with increasing model W concentrations. At the scale of the inner solar system, for most elements, but not all (e.g., Ti, Mo), isotopically distinct mixtures of different nucleosynthetic carrier phases had probably been homogenized by the time at which the chondrite parent bodies formed. However, some early objects other than CAIs may have preserved the isotopic compositions of the least abundant components, as appears to be the case for W isotopes in CV chondrules and in some iron meteorite groups. Finally, nucleosynthetic heterogeneity of Hf isotopes does not significantly affect the 176 Lu–176 Hf and 182 Hf–182 W chronometers in bulk meteorites, but may become relevant for CAIs, particularly when measurement uncertainties will shrink by improved analytical techniques. Acknowledgements We thank Bernard Marty for the editorial handling of this manuscript, and Maria Schönbächler and one anonymous reviewer for constructive comments and suggestions that helped to improve the manuscript. Meteorite samples were provided by Renate Schumacher from the mineralogical museum in Bonn, and by Erik Strub (Allende) from University of Cologne. We thank Toni Schulz and Maike Becker for discussions and for providing the silicate inclusion of Campo del Cielo (T.S.) and the separated CAI of Bali (M.B.). This study was financially supported through DFG grant MU1406/10-1/2. S.P. acknowledges support from Göttingen University through grant 6125/3917543. References Akram, W., Schönbächler, M., Sprung, P., Vogel, N., 2013. Zirconium—hafnium isotope evidence from meteorites for the decoupled synthesis of light and heavy neutron-rich nuclei. Astrophys. J. 777 (2), 169. Arlandini, C., Käppeler, F., Wisshak, K., Gallino, R., Lugaro, M., Busso, M., Straniero, O., 1999. Neutron capture in low-mass asymptotic giant branch stars: cross sections and abundance signatures. Astrophys. J. 525 (2), 886. Bast, R., Scherer, E.E., Sprung, P., Fischer-Gödde, M., Stracke, A., Mezger, K., 2015. A rapid and efficient ion-exchange chromatography for Lu–Hf, Sm–Nd, and Rb– Sr geochronology and the routine isotope analysis of sub-ng amounts of Hf by MC-ICP-MS. J. Anal. At. Spectrom. 30 (11), 2323–2333. Bast, R., Scherer, E.E., Bischoff, A., 2017. The 176 Lu–176 Hf systematics of ALM-A: A sample of the recent Almahata Sitta meteorite fall. Becker, M., Hezel, D.C., Schulz, T., Elfers, B.M., Münker, C., 2015. Formation timescales of CV chondrites from component specific Hf–W systematics. Earth Planet. Sci. Lett. 432, 472–482. Bizzarro, M., Connelly, J.N., Thrane, K., Borg, L.E., 2012. Excess hafnium-176 in meteorites and the early Earth zircon record. Geochem. Geophys. Geosyst. 13 (3). Blichert-Toft, J., Albarède, F., 1997. The Lu–Hf isotope geochemistry of chondrites and the evolution of the mantle–crust system. Earth Planet. Sci. Lett. 148 (1), 243–258. Bouvier, A., Boyet, M., 2016. Primitive Solar System materials and Earth share a common initial 142 Nd abundance. Nature 537 (7620), 399–402. Brennecka, G.A., Borg, L.E., Wadhwa, M., 2013. Evidence for supernova injection into the solar nebula and the decoupling of r-process nucleosynthesis. Proc. Natl. Acad. Sci. USA 110 (43), 17241–17246. Brennecka, G.A., Borg, L.E., Wadhwa, M., 2014. The gadolinium and dysprosium isotopic composition of a supernova injection inferred from Allende CAIs. LPSC abstract #2280. Budde, G., Kleine, T., Kruijer, T.S., Burkhardt, C., Metzler, K., 2016. Tungsten isotopic constraints on the age and origin of chondrules. Proc. Natl. Acad. Sci. USA 113 (11), 201524980.

S.T.M. Peters et al. / Earth and Planetary Science Letters 459 (2017) 70–79

Burkhardt, C., Kleine, T., Bourdon, B., Palme, H., Zipfel, J., Friedrich, J.M., Ebel, D.S., 2008. Hf–W mineral isochron for Ca, Al-rich inclusions: age of the solar system and the timing of core formation in planetesimals. Geochim. Cosmochim. Acta 72 (24), 6177–6197. Burkhardt, C., Kleine, T., Oberli, F., Pack, A., Bourdon, B., Wieler, R., 2011. Molybdenum isotope anomalies in meteorites: constraints on solar nebula evolution and origin of the Earth. Earth Planet. Sci. Lett. 312 (3), 390–400. Burkhardt, C., Kleine, T., Dauphas, N., Wieler, R., 2012. Origin of isotopic heterogeneity in the solar nebula by thermal processing and mixing of nebular dust. Earth Planet. Sci. Lett. 357, 298–307. Burkhardt, C., Schönbächler, M., 2015. Intrinsic W nucleosynthetic isotope variations in carbonaceous chondrites: implications for W nucleosynthesis and nebular vs. parent body processing of presolar materials. Geochim. Cosmochim. Acta 165, 361–375. Cameron, A., Truran, J., 1977. The supernova trigger for formation of the solar system. Icarus 30 (3), 447–461. Charnoz, S., Aléon, J., Chaumard, N., Baillié, K., Taillifet, E., 2015. Growth of calcium– aluminum-rich inclusions by coagulation and fragmentation in a turbulent protoplanetary disk: observations and simulations. Icarus 252, 440–453. Cook, L.D., Kruijer, T.S., Leya, I., Kleine, T., 2014. Cosmogenic 180 W variations in meteorites and re-assessment of a possible 184 Os–180 W decay system. Geochim. Cosmochim. Acta 140, 160–176. Dauphas, N., Cook, L.D., Sacarabany, A., Fröhlich, C., Davis, A.M., Wadhwa, M., Pourmand, A., Rauscher, T., Gallino, R., 2008. 60 Fe in the cosmic blender. Geochim. Cosmochim. Acta, Suppl. 72, 200. 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 in supernova nanoparticles. Astrophys. J. 720 (2), 1577. Fehr, M.A., Rehkämper, M., Halliday, A.N., Hattendorf, B., Guenther, D., 2009. Tellurium isotope compositions of calcium–aluminum-rich inclusions. Meteorit. Planet. Sci. 44 (7), 971–984. Gaffney, A.M., Borg, L.E., 2014. A young solidification age for the lunar magma ocean. Geochim. Cosmochim. Acta 140, 227–240. Gounelle, M., Meynet, G., 2012. Solar system genealogy revealed by extinct shortlived radionuclides in meteorites. Astron. Astrophys. 545. p. A4. Harper, C.L., Jacobsen, S.B., 1996. Evidence for 182 Hf in the early solar system and constraints on the timescale of terrestrial accretion and core formation. Geochim. Cosmochim. Acta 60 (7), 1131–1153. Holst, J.C., Paton, C., Wielandt, D., Bizzarro, M., 2015. Tungsten isotopes in bulk meteorites and their inclusions—implications for processing of presolar components in the solar protoplanetary disk. Meteorit. Planet. Sci. 50 (9), 1643–1660. Hunt, A., Schönbächler, M., 2015. The distribution of p-process 190 Pt in the early solar system, Goldschmidt Abstracts, 1350. Ireland, T.R., Fegley Jr., B., 2000. The solar system’s earliest chemistry: systematics of refractory inclusions. Int. Geol. Rev. 42 (10), 865–894. Ivanova, M.A., Krot, A.N., Nagashima, K., MacPherson, G.J., 2012. Compound ultrarefractory CAI-bearing inclusions from CV3 carbonaceous chondrites. Meteorit. Planet. Sci. 47 (12), 2107–2127. Klay, N., Käppeler, F., Beer, H., Schatz, G., 1991. Nuclear structure of Lu-176 and its astrophysical consequences. II. Lu 176, a thermometer for stellar helium burning. Phys. Rev. C 44 (6), 2839. Kleine, T., Mezger, K., Münker, C., Palme, H., Bischoff, A., 2004. 182 Hf–182 W isotope systematics of chondrites, eucrites, and martian meteorites: chronology of core formation and early mantle differentiation in Vesta and Mars. Geochim. Cosmochim. Acta 68 (13), 2935–2946. Kollar, D., Michel, R., Masarik, J., 2006. Monte Carlo simulation of GCR neutron capture production of cosmogenic nuclides in stony meteorites and lunar surface. Meteorit. Planet. Sci. 41 (3), 375–389. Krot, A.N., Nagashima, K., Wasserburg, G.J., Huss, G.R., Papanastassiou, D., Davis, A.M., Hutcheon, I.D., Bizzarro, M., 2014. Calcium–aluminum-rich inclusions with fractionation and unknown nuclear effects (FUN CAIs): I. Mineralogy, petrology, and oxygen isotopic compositions. Geochim. Cosmochim. Acta 145, 206–247. Kruijer, T.S., Sprung, P., Kleine, T., Leya, I., Burkhardt, C., Wieler, R., 2012. Hf–W chronometry of core formation in planetesimals inferred from weakly irradiated iron meteorites. Geochim. Cosmochim. Acta 99, 287–304. Kruijer, T.S., Kleine, T., Fischer-Gödde, M., Burkhardt, C., Wieler, R., 2014a. Nucleosynthetic W isotope anomalies and the Hf–W chronometry of Ca–Al-rich inclusions. Earth Planet. Sci. Lett. 403, 317–327. Kruijer, T.S., Touboul, M., Fischer-Gödde, M., Bermingham, K.R., Walker, R.J., Kleine, T., 2014b. Protracted core formation and rapid accretion of protoplanets. Science 344 (6188), 1150–1154. Leya, I., Schönbächler, M., Krähenbühl, U., Halliday, A.N., 2009. New titanium isotope data for Allende and Efremovka CAIs. Astrophys. J. 702 (2), 1118. Lugaro, M., Heger, A., Osrin, D., Goriely, S., Zuber, K., Karakas, A.I., Gibson, B.K., Doherty, C.L., Lattanzio, J.C., Ott, U., 2014. Stellar origin of the 182 Hf cosmochronometer and the presolar history of solar system matter. Science 345 (6197), 650–653.

79

Mayer, B., Wittig, N., Humayun, M., Leya, I., 2015. Palladium isotopic evidence for nucleosynthetic and cosmogenic isotope anomalies in IVB iron meteorites. Astrophys. J. 809 (2), 180. McCulloch, M.T., Wasserburg, G.J., 1978. Barium and neodymium isotopic anomalies in the Allende meteorite. Astrophys. J. 220, L15–L19. Mishra, R.K., Chaussidon, M., 2014. Fossil records of high level of 60 Fe in chondrules from unequilibrated chondrites. Earth Planet. Sci. Lett. 398, 90–100. Moynier, F., Fujii, T., Albarède, F., 2009. Nuclear field shift effect as a possible cause of Te isotopic anomalies in the early solar system—an alternative explanation of Fehr et al. (2006 and 2009). Meteorit. Planet. Sci. 44 (11), 1735. Mughabghab, S.F., 2003. Thermal neutron capture cross sections resonance integrals and g-factors. Report, INDC (NDS)-440, IAEA NDS. Münker, C., Weyer, S., Scherer, E., Mezger, K., 2001. Separation of high field strength elements (Nb, Ta, Zr, Hf) and Lu from rock samples for MC-ICPMS measurements. Geochem. Geophys. Geosyst. 2 (12). Patchett, P.J., Tatsumoto, M., 1980. Lu–Hf total-rock isochron for the eucrite meteorites. Nature 288 (5791), 571–574. Patchett, P.J., Tatsumoto, M., 1981. A routine high-precision method for Lu–Hf isotope geochemistry and chronology. Contrib. Mineral. Petrol. 75 (3), 263–267. Peters, S.T.M., Münker, C., Becker, H., Schulz, T., 2014. Alpha-decay of 184 Os revealed by radiogenic 180 W in meteorites: half life determination and viability as geochronometer. Earth Planet. Sci. Lett. 391, 69–76. Peters, S.T.M., Münker, C., Wombacher, F., Elfers, B.M., 2015. Precise determination of low abundance isotopes (174 Hf, 180 W and 190 Pt) in terrestrial materials and meteorites using multiple collector ICP-MS equipped with 1012  Faraday amplifiers. Chem. Geol. 413, 132–145. Pfeifer, M., Lloyd, N.S., Peters, S.T.M., Wombacher, F., Elfers, B.-M., Münker, C., accepted. Tantalum isotope ratio measurements and isotope abundances determined by MC-ICP-MS using amplifiers equipped with 1010 , 1012 and 1013 Ohm resistors. J. Anal. At. Spectrom. http://dx.doi.org/10.1039/c6ja00329j. Qin, L., Dauphas, N., Wadhwa, M., Markowski, A., Gallino, R., Janney, P.E., Bouman, C., 2008. Tungsten nuclear anomalies in planetesimal cores. Astrophys. J. 674 (2), 1234. Qin, L., Nittler, L.R., Alexander, C.O.D., Wang, J., Stadermann, F.J., Carlson, R.W., 2011. Extreme 54 Cr-rich nano-oxides in the CI chondrite Orgueil – implication for a late supernova injection into the solar system. Geochim. Cosmochim. Acta 75 (2), 629–644. Quitté, G., Halliday, A.N., Meyer, B.S., Markowski, A., Latkoczy, C., Günther, D., 2007. Correlated iron-60, nickel-62, and zirconium-96 in refractory inclusions and the origin of the solar system. Astrophys. J. 655 (1), 678. Rauscher, T., Dauphas, N., Dillmann, I., Fröhlich, C., Fülöp, Z., Gyürky, G., 2013. Constraining the astrophysical origin of the p-nuclei through nuclear physics and meteoritic data. Rep. Prog. Phys. 76 (6), 066201. Regelous, M., Elliott, T., Coath, C.D., 2008. Nickel isotope heterogeneity in the early Solar System. Earth Planet. Sci. Lett. 272 (1), 330–338. Sahijpal, S., Goswami, J.N., Davis, A.M., Grossman, L., Lewis, R.S., 1998. A stellar origin for the short-lived nuclides in the early solar system. Nature 391 (6667), 559–561. Schulz, T., Münker, C., Palme, H., Mezger, K., 2009. Hf–W chronometry of the IAB iron meteorite parent body. Earth Planet. Sci. Lett. 280 (1), 185–193. Schulz, T., Münker, C., Peters, S.T.M., 2013. p-Process 180 W anomalies in iron meteorites: nucleosynthetic versus non-nucleosynthetic origins. Earth Planet. Sci. Lett. 362, 246–257. Shollenberger, Q.R., Brennecka, G.A., Borg, L.A., 2016. Clues to the isotopic evolution of the solar system from Er and Yb in Allende CAIs. LPSC abstract #1964. Sprung, P., Scherer, E.E., Upadhyay, D., Leya, I., Mezger, K., 2010. Non-nucleosynthetic heterogeneity in non-radiogenic stable Hf isotopes: implications for early solar system chronology. Earth Planet. Sci. Lett. 295 (1), 1–11. Sprung, P., Kleine, T., Scherer, E.E., 2013. Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth Planet. Sci. Lett. 380, 77–87. Stracke, A., Palme, H., Gellissen, M., Münker, C., Kleine, T., Birbaum, K., Günther, D., Bourdon, B., Zipfel, J., 2012. Refractory element fractionation in the Allende meteorite: implications for solar nebula condensation and the chondritic composition of planetary bodies. Geochim. Cosmochim. Acta 85, 114–141. Tang, H., Dauphas, N., 2012. Abundance, distribution, and origin of 60 Fe in the solar protoplanetary disk. Earth Planet. Sci. Lett. 359, 248–263. Trinquier, A., Elliott, T., Ulfbeck, D., Coath, C., Krot, A.N., Bizzarro, M., 2009. Origin of nucleosynthetic isotope heterogeneity in the solar protoplanetary disk. Science 324 (5925), 374–376. Walker, R.J., 2012. Evidence for homogeneous distribution of osmium in the protosolar nebula. Earth Planet. Sci. Lett. 351, 36–44. Wisshak, K., Voss, F., Kaeppeler, F., Kazakov, L., Beˇcváˇr, F., Krtiˇcka, M., Gallino, R., Pignatari, M., 2006. Fast neutron capture on the Hf isotopes: cross sections, isomer production, and stellar aspects. Phys. Rev. C 73 (4), 045807.