Extremely Na- and Cl-rich chondrule from the CV3 ...starplan.dk/sites/starplan.dk/files/wasserburg, et al.pdf · Extremely Na- and Cl-rich chondrule from the CV3 carbonaceous chondrite

Available online at www.sciencedirect.com

www.elsevier.com/locate/gca

Geochimica et Cosmochimica Acta 75 (2011) 4752–4770

Extremely Na- and Cl-rich chondrule from the CV3carbonaceous chondrite Allende

G.J. Wasserburg a, I.D. Hutcheon b, J. Aleon c, E.C. Ramon b, A.N. Krot d,⇑,K. Nagashima d, A.J. Brearley e

a The Lunatic Asylum, California Institute of Technology, MC 170-25, Pasadena, CA 91125, USAb Glenn Seaborg Institute, Lawrence Livermore National Laboratory, Livermore, CA 94551, USA

c CSNSM IN2P3-CNRS, Batiment 104, 91405 Orsay Campus, Franced Hawai’i Institute of Geophysics and Planetology, School of Ocean, Earth Science and Technology, University of Hawai’i at Manoa,

Honolulu, HI 96822, USAe University of New Mexico, Albuquerque, NM 87131, USA

Received 28 November 2010; accepted in revised form 10 May 2011; available online 17 June 2011

Dedicated to the memory of Brian Mason

Abstract

We report on a study of Al3509, a large Na- and Cl-rich, radially-zoned object from the oxidized CV carbonaceouschondrite Allende. Al3509 consists of fine-grained ferroan olivine, ferroan Al-diopside, nepheline, sodalite, and andradite,and is crosscut by numerous veins of nepheline, sodalite, and ferroan Al-diopside. Some poorly-characterized phases offine-grained material are also present; these phases contain no significant H2O. The minerals listed above are commonlyfound in Allende CAIs and chondrules and are attributed to late-stage iron-alkali-halogen metasomatic alteration of pri-mary high-temperature minerals. Textural observations indicate that Al3509 is an igneous object. However, no residualcrystals that might be relicts of pre-existing CAI or chondrule minerals were identified. To establish the levels of 26Aland 36Cl originally present, 26Al–26Mg and 36Cl–36S isotopic systematics in sodalite were investigated. Al3509 showsno evidence of radiogenic 26Mg*, establishing an upper limit of the initial 26Al/27Al ratio of 3 � 10�6. All sodalite grainsmeasured show large but variable excesses of 36S, which, however, do not correlate with 35Cl/34S ratio. If these excessesare due to decay of 36Cl, local redistribution of radiogenic 36S* after 36Cl had decayed is required. The oxygen-isotopepattern in Al3509 is the same as found in secondary minerals resulting from iron-alkali-halogen metasomatic alteration ofAllende CAIs and chondrules and in melilite and anorthite of most CAIs in Allende. The oxygen-isotope data suggestthat the secondary minerals precipitated from or equilibrated with a fluid of similar oxygen-isotope composition. Theseobservations suggest that the formation of Al3509 and alteration products in CAIs and chondrules in Allende requires avery similar fluid phase, greatly enriched in volatiles (e.g., Na and Cl) and with D17O � �3&. We infer that internalheating of planetesimals by 26Al would efficiently transfer volatiles to their outer portions and enhance the formationof volatile-enriched minerals there. We conclude that the site for the production of Na- and Cl-rich fluids responsiblefor the formation of Al3509 and the alteration of the Allende CAIs and chondrules must have been on a protoplanetarybody prior to incorporation into the Allende meteorite. Galactic cosmic rays cannot be the source of the inferred initial36Cl in Allende. The problem of 36Cl production by solar energetic particle (SEP) bombardment and the possibility that36Cl and 41Ca might be the product of neutron capture resulting from SEP bombardment of protoplanetary surfaces are

0016-7037/$ - see front matter � 2011 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gca.2011.06.004

⇑ Corresponding author. Tel.: +1 808 956 3900; fax: +1 808 956 6322.E-mail address: [email protected] (A.N. Krot).

http://dx.doi.org/10.1016/j.gca.2011.06.004

mailto:[email protected]

http://dx.doi.org/016/j.gca.2011.06.004

Extremely Na- and Cl-rich chondrule from Allende 4753

discussed. This hypothesis can be tested comparing inferred “initial” 36Cl with neutron fluencies measured on the samesamples and on phases showing 36S* by Sm and Gd isotopic measurements.� 2011 Elsevier Ltd. All rights reserved.

1. INTRODUCTION

We report on the mineralogy, petrology, chemistry, oxy-gen isotopes, 26Al–26Mg and 36Cl–36S isotope systematics ofa large Al-rich object, Al3509, from the CV carbonaceouschondrite Allende. This object, which by meteorite standardsis extremely rich in Na (�10 wt%), Cl (�1 wt%), and otherhalogens, was discovered by Clarke et al. (1970). The majorphases were identified as ferroan olivine (Fa26), nepheline,sodalite, and ferroan Al-diopside (Mason and Taylor,1982). Mason and Martin (1977) determined minor and traceelement abundances of Al3509; the concentrations of rareearth elements are at about chondritic level, but with a posi-tive Eu anomaly (Eu/Eu* = 2.9). Wasserburg and Huneke(1979) showed that Al3509 is extremely enriched in 129Xe,produced by decay of 129I (t1/2 � 15.7 Myr), with an initial129I/127I ratio of 1.1 � 10�4, in accord with the levelsfirst reported by Reynolds (1963) in typical chondrites(Wasserburg, 1985).

The high enrichment of halogens in Al3509 makes itimportant in searching for excess radiogenic 36S from thedecay of 36Cl (t1/2 � 0.3 Myr). Excesses of 36S correlatedwith 35Cl/34S ratios have been recently reported in second-ary sodalite (Na8Al6Si6O24Cl2) replacing melilite, plagio-clase or glassy mesostasis in calcium–aluminum-richinclusions (CAIs) and chondrules from the Allende andNingqiang (CV anomalous) meteorites and provide strongevidence for the in situ decay of 36Cl (Lin et al., 2005;Hsu et al., 2006; Ushikubo et al., 2007; Nakashima et al.,2008) after 26Al had decayed. While these studies demon-strate unequivocally that 36Cl was extant in the early solarsystem, they do not explain the absence of large correlatedeffects in 36Ar that would be expected, as 98.1% of 36Cl de-cays to 36Ar (e.g., Villa et al., 1981; Goebel et al., 1982).Existing data also fail to constrain the initial abundanceof 36Cl, because not all sodalite grains within a CAI or achondrule from CV chondrites are found to show 36S ex-cesses correlated with 35Cl/34S ratios; some sodalite grainsshow no resolvable 36S excess (Nakashima et al., 2008).The highest value of the inferred initial 36Cl/35Cl ratio re-ported in CV sodalite is �4 � 10�6. Recently, Jacobsenet al. (2009) and Matzel et al. (2010) reported evidencefor a much higher initial abundance of 36Cl in wadalite[Ca6(Al,Si,Mg)7O16Cl3] in the Allende Type B CAI AJEF.For Type B CAIs, the inferred initial 36Cl/35Cl in wadaliteranges from �2 � 10�5 in AJEF to �3 � 10�7 in Egg-6 andTS34. In these CAIs, wadalite, grossular, monticellite, Al-rich, Ti-poor pyroxene, wollastonite, and Na-rich melilitereplace akermanitic melilite and anorthite in the cores ofthe inclusions, whereas sodalite together with nepheline,secondary anorthite, and ferroan olivine replaces gehleniticmelilite and anorthite in both CAI mantles and Wark–Lov-ering rims (Ishii et al., 2010; Krot et al., 2010). The natureof the widespread volatile-rich alteration process remains

controversial; two classes of models, asteroidal (Krotet al., 1998a,b; Choi et al., 2000; Jogo et al., 2009) and neb-ular (Kimura and Ikeda, 1995; Weisberg and Prinz, 1998),have been discussed. According to the asteroidal models,CV chondrites experienced fluid-assisted thermal metamor-phism under variable conditions (temperature and water/rock ratio) on the CV parent asteroid, which resulted inmobilization of Ca, Si, Fe, Mg, Mn, Na, Cl, and S, the for-mation of a volatile-rich fluid and replacement of primaryphases in chondrules, CAIs and in the matrices. Accordingto the nebular models, chondrules and refractory inclusionsin oxidized CV chondrites were exposed to a highly oxidiz-ing nebular gas resulting in replacement of primary miner-als by ferroan olivine, salite-hedenbergite pyroxenes,andradite, sodalite, and nepheline.

Observations show that 36Cl–36S and 26Al–26Mg system-atics of sodalite in CV CAIs are not correlated indicatingthat the production mechanisms of 36Cl and 26Al are differ-ent (Lin et al., 2005; Hsu et al., 2006). Theoretical calcula-tions of the production of 36Cl in both asymptotic giantbranch (AGB) star and supernova (SN) sources suggestthat these stellar sources are insufficient to produce the re-ported 36Cl/35Cl ratios. Thus, it was inferred 36Cl must beproduced by energetic charged particle bombardment(SEP) from the early Sun (Hsu et al., 2006). Based on themineralogy, petrography, O-isotope compositions,26Al–26Mg and 36Cl–36S isotope systematics of the second-ary mineralization, short-lived 36Cl is inferred to have beenpresent in late formed volatile-rich zones of protoplanets(Jacobsen et al., 2009; Matzel et al., 2010; Krot et al.,2010). The specific means of production of 36Cl and the siteof the irradiation have not yet been identified.

2. ANALYTICAL TECHNIQUES

Two polished thin sections, Mann#1 and Mann#2, weremapped in Mg, Ca, Al, Si, Na, Cl, Cr, and Ti Ka X-raysusing a fully focused electron beam, 15 kV acceleratingvoltage, 100 nA beam current, 20 ms per pixel acquisitiontime, and resolution of �5 lm per pixel with wavelengthdispersive spectrometers on a Cameca SX-50 electronmicroprobe at the University of Hawai’i (UH). The elemen-tal maps in Mg, Ca and Al Ka and Cl, Na, and Fe Ka werecombined using an RGB-color scheme (Mg, Cl – red, Ca,Na – green, Al, Fe – blue) and ENVI (ENvironment forVisualizing Images, ITT Corporation) software package.The mineralogy of the chondrule was studied in backscat-tered electrons (BSE) using a JEOL JXA-8500F field-emis-sion electron microprobe. Electron probe microanalyseswere performed with the JEOL JXA-8500F microprobeusing a 15 kV accelerating voltage, 15 nA beam current,beam size of 1 and 5 lm and wavelength dispersive X-rayspectroscopy. For each element, counting times on both

4754 G.J. Wasserburg et al. / Geochimica et Cosmochimica Acta 75 (2011) 4752–4770

peak and background were 30 s. Matrix effects were cor-rected using PAP procedures (Pouchou and Pichoir, 1984).

Oxygen isotopic compositions were analyzed in situ withthe UH Cameca ims-1280 ion microprobe using two analyt-ical protocols. In the both protocols, the secondary ionmass spectrometer was operated at –10 keV with a 50 eVenergy window and a normal-incidence electron flood gunwas used for charge compensation on the sample. To mea-sure O-isotope compositions of coarse ferroan Al-diopside,a 2 nA focused Cs+ primary ion beam rastering over10 � 10 lm2 area was used. Three oxygen isotopes weremeasured simultaneously 16O– was measured on a multicol-lector Faraday cup (FC) with a 1010 ohm resistor; 17O– wasmeasured with the monocollector electron multiplier (EM);and 18O– was measured on a multicollector FC with a 1011

ohm resistor. To measure O-isotope compositions of fine-grained (<10 lm) ferroan olivine, nepheline, sodalite, andunidentified Fe, Mg, Al-silicate, a �30 pA Cs+ primaryion beam focused to �2 lm was used for both pre-sputter-ing and data collection. 16O– was measured on a multicol-lector FC with 1011 ohm resistor; 17O– was measured withthe monocollector EM, and 18O– was measured with a mul-ticollector EM. Total measurement time per spot was about15 min including pre-sputtering. In both cases, the massresolving power (m/Dm) for 16O– and 18O– was �2000,and that for 17O– was �5500–6000, sufficient to separateinterfering 16OH–. The 16OH– signal intensity was moni-tored after every analysis to estimate the contribution of16OH– interference to the 17O– signal. No evidence of OHhotspots was identified in any areas analyzed. The contribu-tions in all spots were smaller than 0.1& in d17O and nocorrection was made. Data were corrected for instrumentalfractionation using San Carlos olivine, Cr-augite, andMiyakejima anorthite standards. For minerals for whichappropriate standards were lacking, the anorthite standardwas used for nepheline and sodalite, and the olivine wasused for unidentified Fe, Mg, Al-silicate. These instrumen-tal fractionation corrections are mass-dependent and donot affect D17O (D17O = d17O – 0.52 � d18O) values. The re-ported uncertainties include both the internal measurementprecision on an individual analysis and the external repro-ducibility for standard measurements during a given analyt-ical session. For the first protocol, the point-to-pointreproducibility (external reproducibility) on the multipleanalyses of the standards was 0.5�1& (2 times the standarddeviation of the mean, 2r) for both d17O and d18O. For thesecond protocol, the external reproducibility on the multi-ple analyses of the standards was 1�2& and 1.5�2.5&

(2r) for d17O and d18O, respectively. To verify the positionsof the sputtered region, the minerals studied for oxygen iso-topes were photographed in secondary and backscatteredelectrons using the JEOL 5900LV scanning electron micro-scope before and after ion probe measurements.

Cl–S isotope measurements in sodalite were performedusing the Lawrence Livermore National Laboratory Came-ca NanoSIMS 50 using a primary Cs+ beam of �8 pA anddiameter of �200 nm. Depending upon the size of the soda-lite grains, the primary beam was set to raster over areas be-tween 8 � 8 and 30 � 30 lm2. Because the collectorconfiguration on the Cameca NanoSIMS 50 does not allow

for simultaneous measurements of 35Cl and 36S without sig-nificantly distorting their respective peak shapes, measure-ments were performed in combined peak jumping, multi-collection mode, simultaneously measuring 28Si�, 32S�,34S� and 36S�, and subsequently stepping the magnetic fieldto measure 37Cl�. A mass resolving power of �3600 wasused, sufficient to eliminate any contribution from 12C3

�

or 35ClH� to the 36S� signal. However, this resolution wasnot high enough to resolve potential interferences from dou-bly-charged 35Cl37Cl2� and 56Fe16O2� species. To evaluatesuch effects, we analyzed terrestrial sodalite and wadalite.All the terrestrial samples yield 36S/34S ratios indistinguish-able from normal, irrespective of the 35Cl/34S ratio, effec-tively ruling out any potential 35Cl37Cl2� or 56Fe16O2�

interferences on 36S�. We also carefully monitored and eval-uated the potential interferences from 24Mg48Ca2�,28Si44Ca2�, 40Ca16O2

2� etc., and concluded that these areunlikely sources of error in our 36S� measurement. Due tothe low intensity of 36S�, the background at mass 36 wascarefully evaluated; the mean background intensity forsodalite is �0.013 cps. Data were processed as quantitativeisotopic ratio images by using custom software (L’image;L. Nittler) and were corrected for detector dead time andimage shift from layer to layer. Regions of interest (ROIs)were defined based on elemental composition correlatedwith SEM images, and the isotopic composition for eachROI was calculated by averaging over the replicate layers.

Measured 37Cl�/34S� ion ratios were converted to atom-ic ratios using a relative sensitivity factor of 0.71 ± 0.04 forCl/S, determined from measurements of terrestrial scapo-lite, (Na, Ca)4(Al3Si9O24)Cl. This relative sensitivity factoris similar to 0.69 reported by Nakashima et al. (2008),and the literature values for different minerals which are be-tween 0.58 and 0.83 (Lin et al., 2005; Hsu et al., 2006; Ush-ikubo et al., 2007). Because 36Cl decays by b-decay (98.1%)to 36Ar and by electron capture and positron emission(1.9%) to 36S, the 36Cl/35Cl ratios from the 36S/34S and35Cl/34S measurements were calculated using a fraction of0.019 of all 36Cl decays going to 36S.

The Al–Mg isotope measurements were performed usinga modified ims-3f secondary ion mass spectrometer at theLawrence Livermore National Laboratory following estab-lished procedures (e.g., Kennedy and Hutcheon, 1994;Goswami et al., 1994). The Al-rich phases were locatedby optical and ion imaging and a small field aperture in-serted in the sample image plane to ensure acceptance ofsecondary ions for isotopic analysis only from a singlephase of interest. Replicate measurements were performedin most cases. We also analyzed phases with low Al/Mg ra-tios (ferroan Al-diopside) as well as appropriate terrestrialmaterials (e.g., Burma spinel, Dekalb diopside, San Carlosolivine, plagioclase and fassaite glass, and Lake County andMiakejima plagioclase) at the beginning and at the end ofeach measurement session.

No special effort was made to characterize intrinsicmass-dependent fractionation in Al3509 and we do not re-port data for mass fractionation. The deviations in the mea-sured 26Mg/24Mg ratios from the reference value, denotedby d26Mg, were obtained by assuming a power-law mass-dependent fractionation corrections and using the relation:

g


d26Mg ¼ D26Mg

� ðexpfln½ð25Mg=24MgÞmeas=ð25Mg=24MgÞref �=0:514

� 1Þ � 1000;

whereDiMg = [{(iMg/24Mg)meas/(iMg/24Mg)ref} � 1] �

1000 [per mil]; (i = 25, 26). Reference values for 25Mg/24Mgand 26Mg/24Mg are 0.12663 and 0.13932, respectively(Catanzaro et al., 1966).

Two TEM samples of Al3509 were prepared for studyusing the focused ion beam technique. A FEI Quanta 3DField Emission Gun (FEG) SEM/Focus Ion Beam (FIB)instrument was used. These FIB sections sample a phasethat could not be identified using SEM EDS or EPMAtechniques. A platinum protective layer (2 lm thickness)was deposited in a strip across the area of interest priorto FIB sample preparation to minimize ion beam damage.The samples were removed from the thin section by thein situ lift out technique using an Omniprobe 200 microma-nipulator and were transferred to a Cu TEM half grid. Fi-nal ion milling to electron transparency was carried outwith the samples attached to the TEM grid.

The TEM studies were carried out using a JEOL 2010HRTEM and a JEOL 2010F FEG TEM/STEM FASTEMinstrument equipped with a GATAN GIF 2000 image fil-tering (GIF) system. HRTEM, electron diffraction andEDS X-ray analyses were carried out on the JEOL 2010operating at 200 kV. X-ray analyses were obtained withan Oxford ISIS 200 EDS system using an Oxford PentafetUTW EDS detector. The Cliff–Lorimer thin film approxi-mation was used for quantification of EDS data usingexperimentally-determined k-factors. The JEOL 2010Fwas operated at 197 kV for energy filtered TEM (EFTEM),EFTEM spectral imaging (EFTEMSI), electron energy lossspectroscopy (EELS), high angle annular dark field scan-ning TEM (HAADF-STEM) and HRTEM imaging.

3. RESULTS

3.1. Mineralogy, petrography, and chemical compositions

The Al3509 is a large rounded object. The diameter of�8 mm was reported by Mason and Taylor (1982). In twopolished thin sections available for our study, Mann#1and Mann#2, the diameters of the exposed regions of theobject are 5 and 2.5 mm, respectively. X-ray elemental map-ping revealed a core–mantle structure and a complex radialchemical zoning of the object (Figs. 1–4). Inspection ofFig. 1a and b show the absence of any Mg-rich phases inthe object. This is confirmed at much higher levels of magni-fication. Based on the chemical compositions, several zonesor layers in the core (1–4) and the mantle (5–9) of Al3509 canbe identified; these zones/layers are indicated by numbers inFigs. 1–5. In the polished section Mann#2, only four mantlelayers, 6–9, are exposed. Defocused beam (� 5 lm) electronmicroprobe analyses of fine-grained zones are listed inTable 1. Electron microprobe analyses of nepheline, sodalite,ferroan olivine and ferroan Al-diopside grains in therelatively coarse zones are listed in Table 2.

The object as a whole is Ca- and Al (but not Ti)-rich(Ca + Al2O3 > 10 wt%; Bischoff and Keil, 1984); however,there are large variations in Ca and Al abundances be-tween the core and the inner mantle (5, 6), and betweenthe inner mantle and the outer (8, 9) mantle zones (Table 1;Fig. 1c and d). Although Al3509 is apparently an igneousobject (based on rounded shape and spherulitic texture ofits core, see below), it shows widely variable textures thatappear to represent different crystallization growth pro-cesses or sequences. Al3509 contains none of the Al-richor Mg-rich minerals commonly present in igneous CAIs(compact Type A, Type B, or Type C) or Al-rich chond-rules such as melilite, spinel, Al, Ti-diopside, perovskite,plagioclase, forsterite, or magnesium-rich pyroxenes. In-stead it consists entirely of fine-grained minerals (ferroanolivine, ferroan Al-diopside, nepheline, sodalite, andradite,and an unidentified phase). In Allende CAIs and chond-rules, these minerals are considered to be the product ofiron-alkali-halogen metasomatic alteration of primaryhigh-temperature minerals (Krot et al., 1995, 1998a,b). Be-low we describe the mineralogy and petrography of com-positionally distinct zones/layers in the core and themantle of Al3509.

3.1.1. Core

Zone 1 constitutes the major portion of the core. It isrich in Ca, Na, and Cl (Table 1, and Figs. 1c and 3c andd) and has a spherulitic texture (Fig. 5a–d). The spherulitesare mineralogically-zoned and have a sodalite + ferroan Al-diopside central portion and a sodalite + ferroan olivineperiphery (Fig. 5c and d).

Zone 2 is Cl-poor and has a lower Na content than zone1 (Table 1, and Fig. 3c and d). Calcium content is highlyvariable; Ca-rich and Ca-poor domains can be identified(Fig. 1c). The Ca-rich domains are composed of relativelycoarse anhedral grains of ferroan Al-diopside intergrownwith a very fine-grained Na-rich mineral (Fig. 5b and d)identified tentatively as nepheline using EDS. The Ca-poordomains are very fine-grained and consist of elongated crys-tals separated by a phase with a low mean atomic number(Fig. 5a). The defocused beam electron microprobe analy-ses of the Ca-poor domains have low totals (�85 wt%)due to high porosity and possibly the presence of hydratedminerals.

Zone 3 forms a thin layer around zone 2, has high Naand Cl contents, like zone 1, but lacks spherulitic texturesand is Ca-poor (Table 1 and Figs. 1c and 3c and d). Itmainly consists of sodalite and ferroan olivine.

Zone 4 is represented by coarse elongated grains of fer-roan Al-diopside (Table 2 and Figs. 1 and 5a and b). Thecentral part of the pyroxene grains, indicated by arrowsin Fig. 5b, is partly filled by another mineral phase, whichis too small for identification using EDS. The observationsin transmitted polarized light suggest that the pyroxenegrains crystallized away from the center and most likelyrepresent remnants of veins crosscutting the core ofAl3509. The presence of ferroan Al-diopside in nepheline-sodalite-bearing veins crosscutting the mantle of Al3509

(Figs. 2c and 4a and f) supports this interpretation.

Fig. 1. Maps using Ka X-ray emission of Mg (b), Ca (c), and Al (d), and a combined elemental map in Mg (red), Ca (green), and Al (blue) (a)of Al3509, Mann#1. The maps show that Al3509 is rich in Al and Ca, but contains no identifiable Mg-rich phases, typically observed inCa, Al-rich chondrules and CAIs (e.g., spinel, fassaite, forsteritic olivine, or low-Ca pyroxene). It shows radial chemical zoning (see c and d)and is surrounded by a thin fine-grained rim (FGR). The chemically distinct zones are numbered from 1 to 9 (see text for details). Regionoutlined in (a) is shown in Fig. 5a. (For interpretation of the references to color in this figure legend, the reader is referred to the web version ofthis article.)


3.1.2. Mantle

The core is surrounded by a multilayered mantle(Figs. 1–5). The innermost layer 5, exposed only in sectionMann#1, is thin and discontinuous (Figs. 1 and 3) and con-sists of a very fine-grained material (Fig. 5e) that gives thelow analytical totals (�89 wt%) of the defocused beam elec-tron microprobe analyses (Table 1). This could be due tothe high porosity of the fine-grained material and possiblyto the presence of hydrated minerals.

Layer 6 is composed of compact regions of ferroan oliv-ine-nepheline intergrowths and a fine-grained, porous inter-stitial material rich in Na and Al, possibly nepheline(Fig. 5e). The fine-grained material gives low analytical to-tals (�85 wt%) of electron microprobe analyses, most likelydue to high porosity (Table 1).

The inner (5, 6) and outer (8, 9) mantle layers are sepa-rated by a thin discontinuous layer of andradite (zone 7;Fig. 5e) and crosscut by numerous veins composed of neph-eline, sodalite, and ferroan Al-diopside (Figs. 3 and 4).

Layer 8 consists of ferroan olivine, nepheline, sodalite,ferroan Al-diopside, and abundant rounded inclusions ofa Fe, Mg, Al-silicate phase (Table 2). This phase describedin the following section is commonly surrounded by a single

layer of nepheline or a double layer of nepheline + uniden-tified phase having a needle-shaped morphology (Fig. 5e–g).Occasionally, the Fe, Mg, Al-phase contains needle-likeinclusions of a Na-rich phase, possibly nepheline (Fig. 5f).

The outermost layer 9 consists of hopper and skeletalcrystals of ferroan olivine intergrown with nepheline andsodalite, and minor ferroan Al-diopside (Table 2 andFig. 5g and h). In contrast, most CAIs in CV chondritesare surrounded by multilayered, refractory rims (Warkand Lovering, 1977). These Wark–Lovering rim layers aretypically monomineralic and form a sequence (from insideto outside): spinel ± hibonite ± perovskite, melilite, Al-diopside, and forsterite. In the Allende-like oxidized CVchondrites, the Wark–Lovering rims show clear evidenceof later alteration as the spinel layer is enriched in FeO;the melilite layer is replaced by nepheline + soda-lite + anorthite; and the forsterite rim is enriched in FeOand overgrown by a layer of hedenbergite ± wollaston-ite ± andradite. No evidence of Wark–Lovering rim layersas defined above has been detected in any part of or withinAl3509.

In addition to the mantle layers described above,Mann#2 contains a chemically and texturally distinct

Fig. 2. Maps using Ka X-ray emission of Mg (b), Ca (c), and Al (d), and a combined elemental map in Mg (red), Ca (green), and Al (blue) (a)of Al3509, Mann#2. This section is close to the end of the object and exposes only chemically distinct mantle zones 6–9. Although the samecolor scheme is used in Fig. 1a, there are clear differences in color of Al3509. These differences are an artifact of image processing using ENVIdue to higher proportion of Type I chondrules in Fig. 1. (For interpretation of the references to color in this figure legend, the reader isreferred to the web version of this article.)


region in the outer mantle (Figs. 2 and 4). This region is en-riched in Si (Fig. 4b), depleted in Fe compared to the rest ofthe object, has a quenched, radial texture (Fig. 4c–e), andconsists of skeletal crystals of sub-Ca pyroxene replacedto various degrees by ferroan high-Ca pyroxene, and inter-stitial sodalite. In the peripheral part of the region, theseminerals are replaced by coarse ferroan olivine and sodalite(Fig. 4d).

3.1.3. TEM study of the unidentified Fe, Mg, Al-silicate

phase

To seek an identification of the unidentified a Fe, M-g, Al-silicate phase referred to in the section on layer 8,two FIB sections were extracted from two separate grainsof the phase and were characterized by TEM. Based onelectron diffraction and EDS data, we have identified theunknown phase as ferroan olivine intergrown on a nano-meter scale with an aluminous Fe-rich silicate phase.The TEM observations also confirm that the phase witha needle-shaped morphology is nepheline. Dark fieldSTEM images showing the complex microstructure ofthe olivine in one of the FIB sections are shown inFig. 6a and b. The grains contain discontinuous lamellaeof a second phase with lower Z contrast which typically

occur at the subgrain boundaries between the threedistinct grains. This phase damages rapidly under the elec-tron beam, but is compositionally consistent withnepheline.

High resolution STEM imaging of the high Z phaseshows that very thin lamellae of a higher Z-phase are coher-ently intergrown within the host phase. These lamellae havewidths ranging from 1 to 2 nm. The olivine grains showconsiderable evidence of internal strain (Fig. 6d), indicatedby the complex mottled contrast in the image and strongasterism in diffraction maxima in the electron diffractionpattern (Fig. 6b). In addition to evidence of significantstrain, the olivines contain distinct nanometer-sized (10–100 nm) voids that are distributed heterogeneously throughthe grains (Fig. 6a, b, e and f).

X-ray analysis of the olivine and high-Z lamellae werecarried out using STEM EDS techniques. The host olivinehas a composition of Fa3840. The high-Z lamellae are toothin to determine their compositions precisely, but theyare clearly enriched in Al2O3 (up to 22 wt%). However,all the analyses are mixtures of the olivine and lamellaphase, demonstrating that the lamella are extremely alumi-nous, consistent with the EPMA data (Table 2). Somelamellae are also enriched in Fe, but the data are quite

Fig. 3. Maps using Ka X-ray emission of Fe (b), Na (c), and Cl (d), and a combined elemental map in Cl (red), Na (green), and Fe (blue) (a) ofAl3509, Mann#1. The object is low in Fe and highly enriched in Cl and Na compared to the Allende matrix. The yellow and green colors in (a)correspond to sodalite and nepheline, respectively. The chemically distinct zones are numbered from 1 to 9 (see text for details). Regionoutlined in (a) is shown in Fig. 5a. (For interpretation of the references to color in this figure legend, the reader is referred to the web version ofthis article.)


variable with Fe/(Fe + Mg) ratios that overlap that of thehost olivine (Fe/(Fe + Mg) = 0.38–0.48, c.f. olivine = 0.38–0.40). Although it is not possible to identify the lamellaephase unequivocally by EDS analyses alone, the composi-tional data are suggestive of a chlorite mineral. Assumingthe phase is chlorite and calculating a mineral formulabased on 28[O] atoms for the most aluminous analysisyields a mineral formula of (Mg6.1Fe4.7)Al1.8(Si4.3A-l3.7)O20(OH)8. This composition would certainly be Si-defi-cient compared with most chlorite compositions whichtypically contain >5 Si cations per 28 [O]. The X-site sum(octahedral cations, Mg, Fe, Al) is also higher than theideal 12 per formula unit. However, electron diffractiondata provide additional support for the presence of chloriteintergrown with the olivine. Fig. 7 shows an [0 1 1] zoneaxis pattern from a region of olivine grain containing a highdensity of lamellae. Distinct, but diffuse and lower intensityextra reflections parallel to the [1 0 0] direction of the oliv-ine are present. The electron diffraction pattern for theintergrown phase is clearly a zone axis pattern and confirmsthat the lamella phase is coherently intergrown with theolivine. The calculated d-spacing of the intergrown phaseparallel to the [1 0 0] direction of the olivine, �1.42 nm, isconsistent with the (0 0 1) basal spacing of chlorite. How-ever, the d-spacings and angular relationships between the

diffraction vectors are not an exact match for chlorite,although they are close. This discrepancy could be the re-sult of several factors including the fact that the intergrownphase may be significantly distorted due to misfit strainwith the olivine, particularly so because the lamellae areonly one to two unit cells in thickness. In addition, thehighly aluminous composition of the phase may cause amarked change in the lattice parameters compared withthe clinochlore unit cell used to calculate the theoretical dif-fraction data. Based on the compositional and crystallo-graphic data, it therefore seems most probable that thisphase is a highly aluminous chlorite.

3.1.4. Fine-grained rim

The object is surrounded by a fine-grained rim com-posed of ferroan olivine, nepheline, and sodalite (Fig. 8).The rim is depleted in Ca relative to the neighboring matrix.Abundant nodules of Ca, Fe-rich (salite-hedenbergite solidsolution) pyroxenes are observed at boundary between therim and the matrix.

In all of the zones described above and in the veins, thereis no major change in the overall mineralogy, and, hence,no direct evidence of late stage alteration producing a dis-tinctive mineral assemblages except for very minor possiblyhydrous phases.

Fig. 4. Map using Ka X-ray emission of Si (b), a combined elemental map in Cl (red), Na (green), and Fe (blue) (a), and backscatteredelectron images (c–f) of Al3509, Mann#2. The yellow and green colors in (a) correspond to sodalite and nepheline, respectively. This section isclose to the end of the object and exposes only chemically distinct mantle zones 6–9. Regions outlined in (a) (from top to bottom) are shown in(c) and (f), respectively. Regions outlined in (c) (from top to bottom) are shown in (d) and (e), respectively. (b–e) The top region outlined in (a)is enriched in Si compared to the rest of the object and has a quenched, radial texture. It consists of skeletal crystals of sub-Ca pyroxene(identified by EDS only) replaced by ferroan Al-diopside, and interstitial sodalite (sod) (e). In the peripheral part, these minerals arepseudomorphed by ferroan olivine (ol) and sodalite (sod) (d). (f) The mantle zones are crosscut by multiple veins composed of nepheline (nph),sodalite, and ferroan Al-diopside (cpx). (For interpretation of the references to color in this figure legend, the reader is referred to the webversion of this article.)


3.2. Isotopic compositions

3.2.1. Oxygen isotopes

Oxygen isotopes were measured in coarse ferroan Al-diopside grains (veins) in the core of Al3509 (Fig. 9a), inthe Fe, Mg, Al-silicate phase in the mantle layer 8 (Fig. 9band EA1), and in nepheline and sodalite veins crosscuttingthe mantle (Figs. 9c and d, and EA1). To check whetherthe Al3509 minerals contain any significant amount of struc-tural water, the 16OH� count rate in each of the grains ana-lyzed was measured after O-isotope measurements. The16OH� count rates in nepheline, sodalite, ferroan olivine,

and the unidentified phase were similarly low (�104 c/s)and not distinguishable from the background.

Oxygen-isotope data of the grains measured are listed inTable 3. In Fig. 10, the oxygen isotopic compositions ofthese grains are plotted together with the compositions ofthe secondary Ca, Fe-rich pyroxenes (salite-hedenbergitesolid solution) and sodalite in the Allende matrix, in Al-lende dark inclusions (Krot et al., 2000) and in a rimaround the Allende Type B CAI TS24 (Cosarinsky et al.,2008). Oxygen isotopic compositions of all these mineralsplot along a mass-dependent fractionation line at a D17Ovalue of ��2.5 ± 0.5&.

Fig. 5. Backscattered electron images of a region in Al3509, Mann#1 outlined in Fig. 1a. The chemically distinct units (zones) are numberedfrom 1 to 9 in (a). (b–d, e and g) represent a traverse from the core to the edge of the object and illustrate the mineralogy of these zones. (f andh) are from zones 8 and 9 in Mann#2, respectively. Regions outlined in (b) (from top to bottom) are shown in (c and d). and = andradite;cpx = ferroan Al-diopside; fgm = poorly-characterized fine-grained material; nph = nepheline; ol = ferroan olivine; sod = sodalite;un = unidentified mineral phase.


3.2.2. 26Al–26Mg systematics

The Al–Mg isotopic results (2s error bars) for sodaliteand ferroan Al-diopside are listed in Table 4 and shownin Fig. 11. The Al–Mg isotopic results (2r error bars) forsodalite and ferroan Al-diopside are shown in Fig. 11.Al3509 shows no evidence for radiogenic 26Mg* even insodalite with extremely high Al/Mg ratios. The pyroxene

26Mg/24Mg ratio is indistinguishable from the terrestrial va-lue. The upper limit of the inferred initial 26Al/27Al ratio inAl3509 is 3 � 10�6. This is in accord with the upper boundsfound for sodalite samples in alteration products of CAIs(e.g., Lin et al., 2005; Hsu et al., 2006; Ushikubo et al.,2007). Assuming uniform distribution of 26Al in the innersolar system, the low value of (26Al/27Al)0 in sodalite in

Table 1Average defocused beam electron microprobe analyses (in wt%) of fine-grained core and mantle zones in the Al3509 chondrule.

Zones No. an. SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O Cl Total

1, core 11 45.1 0.06 13.6 0.14 5.6 0.08 9.0 14.2 7.7 0.19 2.4 98.10.5 0.02 0.6 0.02 0.9 0.03 1.1 0.8 0.6 0.2 0.2 0.8

1, rim 5 36.1 0.06 27.0 0.16 4.9 0.05 7.9 0.20 17.6 0.02 5.2 99.20.3 0.02 1.4 0.02 1.2 0.02 2.0 0.06 1.6 0.01 0.40 1.0

3 10 34.7 0.09 17.3 0.20 8.8 0.09 13.8 0.82 8.4 0.95 0.12 85.21.3 0.02 0.9 0.02 0.5 0.02 1.2 0.34 0.4 0.05 0.05 2.9

5 8 45.3 0.10 17.8 0.20 2.0 0.01 10.2 8.4 3.9 1.0 0.12 89.01.1 0.02 1.0 0.02 0.1 0.01 0.8 1.5 0.2 0.3 0.03 1.6

6 13 40.94 0.07 17.3 0.15 8.5 0.10 19 4.4 7.8 1.0 0.02 99.21.0 0.03 3.1 0.02 1.6 0.04 3.7 3.0 1.8 0.21 0.02 1.5

No. an. = number of analyses; in italics are 1 standard deviations.

Table 2Electron microprobe analyses (in wt%) of pyroxene, olivine, nepheline, sodalite, and unidentified phase in the Al3509 chondrule.

Mineral SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O K2O Cl Total Fa Fs Wo

px 48.9 0.12 6.7 0.15 3.5 0.04 14.8 23.2 0.62 0.21 n.d. 98.3 – 5.9 50.0px 45.6 0.03 11.0 0.12 5.9 0.04 12.0 24.0 0.31 n.d. n.d. 99.1 – 10.2 52.8px 47.1 0.07 11.6 0.12 4.8 0.09 11.7 23.4 0.70 n.d. n.d. 99.5 – 8.6 53.9nph 43.1 n.d. 35.7 n.d. 0.21 0.04 0.09 0.96 17.4 1.5 n.d. 99.0 – – –ol 37.4 n.d. 0.27 0.10 24.7 0.16 36.6 0.09 0.12 n.d. n.d. 99.5 27.5 – –ol 36.0 0.04 0.30 0.20 31.2 0.27 31.6 0.05 0.10 n.d. n.d. 99.8 35.7 – –ol 35.1 n.d. 0.29 0.09 35.1 0.29 28.9 0.07 0.16 n.d. n.d. 99.9 40.6 – –sod 39.1 n.d. 35.2 n.d. 0.18 n.d. 0.08 0.07 17.9 n.d. 7.1 99.6 – – –un 29.9 0.05 11.1 0.12 30.5 0.25 26.4 0.27 0.19 0.04 0.02 98.9 – – –un 30.9 0.08 8.7 0.11 30.5 0.22 27.0 0.24 0.20 0.04 0.03 98.0 – – –

n.d. = not detected; nph = nepheline; ol = olivine; px = pyroxene; sod = sodalite; un = unidentified phase.


Al3509 relative to the canonical 26Al/27Al ratio of�5.2 � 10�5 reported in whole-rock CV CAIs (Jacobsenet al., 2008) would require a time interval for sodalite for-mation of �3 Myr following the crystallization of CVCAIs.

3.2.3. 36Cl–36S systematics

The Cl–S isotopic results for sodalite are listed in Table 5and plotted in Fig. 12. The Cl–S isotopic results for sodaliteare plotted in Fig. 12. Two features are immediately appar-ent. First, all of the Al3509 sodalite grains measured showlarge excesses of 36S, up to 1.7� the terrestrial value. Sec-ond, the data show no evidence for a correlation betweenthe excess 36S and the 35Cl/34S ratio; the sodalite with thehighest 36S/34S ratio has the third lowest 35Cl/34S ratio(�4850). We have no prima facie evidence for in situ decayof 36Cl in Al3509 from a correlation of 36S with 35Cl. How-ever, the large excesses of 36S in sodalite strongly suggestthat it formed in an environment where 36Cl had once beenpresent. Whether the 36Cl had decayed prior to sodalite for-mation or was incorporated live into the sodalite and sub-jected to later redistribution remains unclear. The lack ofany correlation between the 36S/34S and 35Cl/34S ratios inAl3509 coupled with the extensive evidence for multiplestages of crystallization and recrystallization preclude usingthese data to estimate the initial abundance of 36Cl. Our

preferred interpretation is that abundant 36Cl had beenpresent in this object and that radiogenic 36S was redistrib-uted during the subsequent recrystallization processes.

4. DISCUSSION

4.1. Classification and origin of Al3509

The spherical shape of Al3509, its locally preserved igne-ous, quenched texture (Fig. 5b–d), and the high Ca and Aland low Ti contents show that the object is a Ca, Al-richchondrule. The lack of minerals commonly observed inigneous CAIs (spinel, Al, Ti-diopside, melilite, anorthite,perovskite or ilmenite that commonly replaces perovskitein Allende CAIs (Hashimoto and Grossman, 1987)), andthe lack of the Wark–Lovering rim layers all indicate thatthe object is not an altered CAI. In particular, phases suchas spinel and Al, Ti-diopside are resistant to alteration andsome of them should have been preserved through extensivealteration and metamorphism.

In contrast to the typical Ca, Al-rich chondrules in CVcarbonaceous chondrites, which have porphyritic textures(e.g., Srinivasan et al., 1999; Krot et al., 2002) and pheno-crysts with magnesium-rich, alkali-poor compositions (spi-nel, Al, Ti-diopside, anorthite, olivine, and pyroxenes),Al3509 has a non-porphyritic texture and is very

Fig. 6. STEM and TEM images of FIB-prepared regions of the unidentified Mg–Al–Fe-rich phase (un) shown in Fig. 5. (a) Dark-field STEMZ-contrast image of FIB section of the unknown phase. Based on electron diffraction and EDS analysis the phase is ferroan olivine (Ol) withfine-scale intergrown lamellae of an Al-rich phase, probably chlorite. The image shows three subparallel, elongate grains of olivine (Ol brightgrain) embedded within a porous, inclusion-rich region of nepheline (Ne). The olivine grains show complex variations in Z-contrast due to thepresence of intergrowths of a low-Z phase, nepheline (Ne) and voids which appear as nanometer-sized low-Z features distributedheterogeneously through the olivine. The platinum layer deposited to protect the sample during FIB sample preparation is the high-Z feature(Pt) in the lower part of the image. (b) Higher magnification Z-contrast image of a region at the interface between two of the olivine subgrains.The right hand region of the grain contains abundant, nanometer-wide lamellae of a secondary phase (black arrows) with higher Z-contrast,intergrown with the olivine (Ol). The distribution of the lamellae is heterogeneous within the olivine. Numerous nanometer-sized voids arealso evident in the right hand grain. (c) Bright field TEM image of the olivine (Ol) grain shown in (b), but in a different orientation. Themottled contrast in the image is due to strain within the olivine lattice caused by the presence of subgrains and the intergrown lamellae of thesecondary phase. A region of strain-free olivine on the rim of the grain is indicated by the white arrow. Thick lamellae of nepheline (Ne) whichhave undergone electron beam irradiation to form a porous amorphous material are present. (d) [0 1 1] zone axis electron diffraction patternof olivine showing distinct asterism in electron diffraction maxima due to the strain within the olivine. The asterism is indicative of thepresence of misoriented domains or subgrains within the olivine. Extra diffraction maxima are present (white arrow), between the (h 0 0)diffraction maxima, indicating that the olivine is intergrown with a crystallographically oriented second phase. (e) Bright field TEM imageshowing the olivine in the second FIB section. The olivine (Ol) occurs as several subparallel, elongate grains with different widths intergrownwith lamellae of nepheline (Ne). The large olivine grain in the center contains a heterogeneous distribution of elongate voids. The variablecontrast within the grain is due to the presence of multiple subgrains separated by subgrains that appear to be decorated by subgrainboundary dislocations. (f) Higher magnification bright field image showing a detail of a region of strained olivine (Ol) containing abundantvoids (white arrows). The voids in this region are subrounded and range from a few nanometers to 100 nm in size. Note the inclusion andstrain-free rim on the edge of the olivine grain.


volatile-rich. It consists almost entirely of fine-grained fer-roan olivine, ferroan Al-diopside, nepheline, sodalite,andradite, and a rare Fe, Mg, Al-silicate phase which maybe chlorite. These observations and some of the textures

indicate that either Al3509 rapidly crystallized from a meltrich in Na, Cl, and Fe or that it was a complete replacementof a pr-existing chondrule. In any case, the site of its forma-tion must have occurred in a precursor body to the Allende

Fig. 7. Selected area electron diffraction pattern of the [0 1 1] zoneaxis of olivine showing the presence of extra diffraction maximumbetween the (h 0 0) diffraction rows, caused by the presence of thelamellae of the high-Z phase shown in Fig. 6b. The d-spacing ofthese diffraction maxima is 1.42 nm consistent with the (0 0 1) basalspacing of chlorite. The overall diffraction pattern of the secondphase appears to be most consistent with a [1 0 0] zone axis of achlorite phase. The patterns diffraction patterns show that theolivine and chlorite have a close crystallographic orientationrelationship with [0 1 1]ol//[1 0 0]chl and (1 0 0)ol//(0 0 1)chl.

Fig. 8. Backscattered electron images of a fine-grained rim (FGR)and neighboring Allende matrix (MX) around Al3509, Mann-2chondrule (chd). Region outlined in (a) is shown in (b). The fine-grained rim is mainly composed of ferroan olivine (ol), nepheline(nph), and sodalite (sod). There are abundant rounded nodules ofCa, Fe-rich pyroxenes (salite-hedenbergite solid solution; Ca, Fe-px) in the Allende matrix just outside the fine-grained rim. Theboundaries between the chondrule, fine-grained rim and matrix areindicated by dashed lines.


meteorite. We infer that Al3509 is a Ca, Al-rich chondrulebut not a CAI.

In contrast to typical Ca, Al-rich chondrules found inCV chondrites which have porphyritic textures and magne-sium-rich, alkali-poor phenocryst compositions (Srinivasanet al., 1999; Krot et al., 2002), there are also Ca, Al-richchondrules dominantly composed of glass which have beenfound in unequilibrated ordinary chondrites but not in CVchondrites (Bischoff and Keil, 1984; Krot and Rubin, 1994;Russell et al., 1996; Nehru et al., 2008). These glassychondrules contain some porphyritic crystals or showdevelopment of devitrification with incipient crystals.Glassy chondrules rich in Na are exceptionally rare. Therare examples are Cl-free and always contain either pheno-crysts and/or xenocrysts of Mg-rich minerals (olivine,pyroxene, and/or spinel). Two such chondrules have beenshown to contain evidence for the presence of 26Al(26Al/27Al � 10�5; Russell et al., 1996). In contrast, whileAl3509 shows some evidence that it formed from a moltendroplet, it does not have any other chemical affinity withNa-rich glass chondrules. Nonetheless, the existence ofNa-rich glass chondrules shows that late formation of li-quid silicate droplets rich in the highly volatile elementNa occurred. Due to the high vapor pressure of Na, noneof these Na-rich glass chondrules could have formed underlow nebular pressures at the high temperatures requiredfor melting.

A possible connection between Al3509 and typical(ferromagnesian and Al-rich) chondrules and refractory

inclusions in the Allende meteorite is found in the fact thatthe mineralogy of Al3509, particularly the presence of soda-lite and nepheline, is extremely close to that found in thelate-stage alteration products of the Allende CAIs andchondrules (Ikeda and Kimura, 1995; Kimura and Ikeda,1995, 1998; Krot et al., 2002, 2007, 2010). The nature ofthe metasomatic process that impregnated the AllendeCAIs and chondrules with such very volatile componentsis poorly understood. This event must have occurred insome type of planetary environment prior to the incorpora-tion of the CAIs and chondrules into the Allende parentbody. This environment could be either a “proto-Allende”

or another body/bodies from which Allende subsequentlyformed or was assembled. The presence of extensively meta-somatically altered chondritic lithic clasts (dark inclusions)in the Allende meteorite supports this conclusion (e.g.,Fruland et al., 1978; Kurat et al., 1989; Johnson et al.,1990; Buchanan et al., 1997; Krot et al., 2001; Pravdivtseva

Fig. 9. Backscattered electron images of the ferroan Al-diopside (a), sodalite (b and c), nepheline (d), ferroan olivine (e), and unidentifiedphase (f) measured for oxygen-isotope compositions. In Fig. 7a, regions sputtered during oxygen-isotope measurements are outlined.px = ferroan Al-diopside; nph = nepheline; un = unidentified phase; sod = sodalite.


et al., 2003). This alteration must have taken place in amuch higher density environment than is possible undernebular conditions. The interior of a planetary body or ashock event resulting from a collision between meteoriteparent bodies/proto-asteroids are two plausible scenarios.

The nature of the fluid causing the alteration of theAl3509 precursor is not known. One possibility is thatAl3509 represents the bulk fluid and the observed alterationof CAIs and typical chondrules reflects circulation of thatfluid throughout the proto-Allende body. The rare occur-rences of hydrous minerals in Al3509 argue against a highabundance of H2O in the fluid. This is in accord with theresults of Brearley (1997, 1999), who found evidence ofmodest aqueous alteration but no evidence of a pervasiveH2O-dominated alteration in CAIs or ferromagnesianchondrules in Allende. If water was a significant componentof the bulk fluid, then the phases that precipitated must beanhydrous or they represent the result of dehydrationfrom the original phases that precipitated (Kojima andTomeoka, 1996).

Alternatively, the bulk composition of Al3509 may notrepresent the composition of the fluid and a more mobilecomponent is required. There has been no report of fluidinclusions in CV meteorites, but evidence for the presenceof Na- and Cl-rich volatiles and fluid inclusions with wateras the principal fluid phase has been found in H6 chondritesin the form of halite crystals (Zolensky et al., 1999). A care-ful investigation of this halite showed that it had a 40Ar/39Arage of 4.5 Gyr and contained essentially pure 129Xe from de-cay of 129I with the inferred initial 129I/127I ratio of 1.3 �10�4 (Whitby et al., 2000). Thus, the formation and trans-port of a Na- and Cl-rich fluid must have occurred in proto-planetary bodies during relatively early epochs. Nakashimaet al. (2010) showed that a halite grain from the Zagmeteorite contained no 36S excess (36Cl/35Cl < 3.2 � 10�7).For the case of the Zag halite, the formation and transportof Na- and Cl-rich fluid must have occurred during rela-tively early epochs but after any initial 36Cl had decayed.In the case of Al3509, there is a clear indication that excess36S from decay of 36Cl was present but must have been

Table 3Oxygen isotopic compositions of nepheline (nph), sodalite (sod),ferroan Al-diopside (px), ferroan olivine (ol), and unidentifiedphase (un) in the Al3509 chondrule.

Mineral Point # d17O 2r d18O 2r D17O 2r

px 1 1.5 1.0 8.0 0.7 �2.7 1.0px 2 1.6 1.0 8.0 0.7 �2.6 1.0px 3 2.0 1.0 8.4 0.7 �2.4 1.0px 4 2.1 1.0 8.5 0.6 �2.3 1.0nph 1 �0.8 2.2 3.4 1.9 �2.5 2.5nph 2 �0.3 2.2 3.0 1.9 �1.9 2.4ol 1 �2.9 2.1 0.4 1.2 �3.1 2.2ol 2 �1.8 2.1 1.4 1.1 �2.5 2.2sod 1 �1.9 2.1 �0.1 2.0 �1.8 2.4sod 2 �2.8 2.2 1.0 2.0 �3.4 2.5sod 3 �2.1 2.3 0.4 1.9 �2.3 2.5sod 4 �2.8 2.3 0.8 2.0 �3.3 2.5sod 5 �2.9 2.3 0.5 1.9 �3.1 2.5un 1 0.1 1.9 2.8 1.2 �1.4 2.0un 2 1.0 2.0 2.9 1.2 �0.5 2.1un 3 �0.1 2.1 3.4 1.2 �1.9 2.2un 4 �0.5 2.0 2.9 1.2 �2.0 2.1

Fig. 10. Three-isotope oxygen diagram of ferroan Al-diopside (px),nepheline (nph), sodalite (sod), ferroan olivine (ol) and unidentifiedphase in Al3509 (this study) and of secondary Ca, Fe-richpyroxenes (salite-hedenbergite solid solution; px) and sodalite inthe Allende matrix, rim around Allende CAI TS24 (data fromCosarinsky et al., 2003), and Allende dark inclusion 3529 (datafrom Krot et al., 2000); error bars are 2r. The terrestrialfractionation (TF) line and carbonaceous chondrite anhydrousmineral (CCAM) line are shown for reference. In Fig. 8b, the samedata are plotted as D17O = d17O – 0.52 � d18O. Oxygen isotopiccompositions of all these phases plot along mass-dependentfractionation line with D17O value of �–2.5 ± 0.5& suggestingprecipitation from a fluid of similar O-isotope composition.Locations of grains measured for oxygen isotopic compositionsin Al3509 are shown in Fig. EA1.


subject to recrystallization and redistribution. While 36Arhas not been studied in this object, all previous studies indi-cate that 36Ar would only be present at very low levels cor-responding to 36Cl/37Cl of �10�8 (Villa et al., 1981; Goebelet al., 1982). It appears that this dilemma may be resolvedfrom the work of Turner et al. (2009) who showed that Arcan be effectively lost relative to Xe during heating due tolower activation energy for Ar diffusion. Release of essen-tially all of the Cl-correlated Ar and retention of the I-cor-related Xe on a timescale of 1 Myr would requiresustained temperatures of approximately 460 �C. This inter-pretation would then require that all such samples hadundergone a post-formation heating event. If we assumeno radiogenic 36Ar in Al3509, then, the presence of large ex-cesses of radiogenic 129Xe* require similar post-formationheating. This model is in consonance with the isotopic andtextural observations presented above.

4.2. Origin of 36Cl in Al3509

A distinct problem arises with regard to 36Cl. The originof 36Cl has been attributed to SEP irradiation by an activeearly Sun as no stellar source appears possible (Wasserburget al., 2006). However, the results of dust irradiation modelsdo not appear to yield the abundances of other short-livednuclei observed in meteorites, e.g., 26Al (e.g., Goswamiet al., 2001; Leya et al., 2003; Duprat and Tatischeff,2007). Here we consider the possibility of secondary neu-trons from an SEP source for producing 36Cl. This leadsto other requirements as to the nature of the possible tar-gets and possible observable isotope effects.

Insofar as any bombardment by energetic charged parti-cles (Galactic Cosmic Rays (GCR) or Solar Energetic Par-ticles (SEP) from a T-Tauri Sun) occurs to protoplanetarybodies or planetary bodies, then secondary neutrons willplay an important role as has long been known for themeteorites (cf Eugster et al., 1970, 2006) and the Moon

(e.g., Eugster et al., 1970; Russ et al., 1971). This will resultin the capture of secondary neutrons of epithermal energyon other nuclei with large neutron capture cross sections.This is in addition to more direct spallation products. Thereis a spatial separation of the regions of major neutron fluxfrom that where direct spallation reactions are dominant.

Table 4Aluminum–magnesium isotope systematics of sodalite and pyrox-ene in the Al3509 chondrule.

Mineral 27Al/24Mg ± 2r 26Mg/24Mg ± 2r

Pyroxene 1 0.9 ± 0.1 0.13939 ± 0.00028Pyroxene 3 0.5 ± 0.1 0.13929 ± 0.00020Sodalite 1 337 ± 36 0.1397 ± 0.0025Sodalite 2 869 ± 96 0.1402 ± 0.0035Sodalite 3 267 ± 45 0.1395 ± 0.0013Sodalite 5 1453 ± 120 0.1407 ± 0.0032Sodalite 6 455 ± 49 0.1389 ± 0.0011

Fig. 11. Aluminum–magnesium isotope diagram of sodalite andferroan Al-diopside in Al3509.

Table 5Chlorine–sulfur isotope systematics of sodalite in the Al3509

chondrule.

Mineral 35Cl/34S ± 2r 36S/34S ± 2r

Al 3509 sodalite

#4 (24.1 ± 2.4) � 103 0.00553 ± 0.00043#6 (3.9 ± 0.5) � 103 0.00455 ± 0.00027#7 (1.7 ± 0.2) � 103 0.00446 ± 0.00027#8 (4.9 ± 0.5) � 103 0.00602 ± 0.00038#9 (62.2 ± 5.7) � 103 0.00473 ± 0.00085

Canyon diablo troilite

CDT1 0.0022 ± 0.0006 0.003514 ± 0.000016

Terrestrial sodalite

Bahia 1 (13.8 ± 1.2) � 103 0.00325 ± 0.00040Bahia 13 (10.4 ± 0.9) � 103 0.00354 ± 0.00034Bahia 16 (13.3 ± 1.4) � 103 0.00287 ± 0.00064Bahia 19 (2.1 ± 0.2) � 103 0.00304 ± 0.00019Bahia 20 (2.2 ± 0.2) � 103 0.00339 ± 0.00016Bahia 21 (12.8 ± 1.2) � 103 0.00375 ± 0.00084Bahia 22 (16.5 ± 1.5) � 103 0.00405 ± 0.00035Brevig 1 (1.3 ± 0.1) � 103 0.00333 ± 0.00008Brevig 2 (1.2 ± 0.1) � 103 0.00344 ± 0.00011Brevig 3 (1.2 ± 0.1) � 103 0.00340 ± 0.00012

Fig. 12. Chlorine–sulfur isotope diagram of sodalite in Al3509 andof terrestrial sodalite.

Fig. 13. A schematic illustration of evolution of 36S/34S ratio for aseries of phases with different initial 35Cl/34S ratios that weresubject to a total epithermal neutron fluence of W, where r is theneutron capture cross section and f is the fraction of decays to 36S.This is analogous to a standard “isochron” evolution diagram.However, the inferred “initial” 36Cl/35Cl is really the total numberof 36Cl produced by neutron capture on 35Cl over the total neutronexposures of the assemblage.


The consequences of neutron capture on 35Cl for producing36Cl are schematically illustrated in Fig. 13.

Significant neutron capture effects in Allende CAIs(including neutron capture on 35Cl producing 36Ar) wereearly reported by Goebel et al. (1982), but were attributed

to GCR and not to SEP from the early Sun. Clear evidenceof 36Cl production by neutron capture was found in Al-lende by Nishiizumi et al. (1986) by counting cosmic rayproduced nuclei as a function of depth. At 25 cm depth theyfound 140 dpm/m of 36Cl produced by neutron capture.The peak of neutron production would occur at�200 gm/cm2, a factor of two higher. For a cosmic rayexposure age of �5 Myr (Fireman et al., 1970; Eugsteret al., 2006), this would correspond to a total neutron flu-ence at the peak of 6.8 � 1014 n/cm2. If we consider thefractional depletion of 149Sm found for Allende of4 � 10�5 and the corresponding enrichment of 150Sm re-ported by Carlson et al. (2007), then using the cross sectionof r149 � 4 � 10�20 cm2, we obtain a total neutron fluenceof W = 1015 n/cm2 in good agreement with the above calcu-lation. This calculation is in agreement with earlier studies


by Hidaka et al. (2000) and Hidaka and Yoneda (2009) whodid extensive measurements on Sm and Gd. They inferredthat for bulk Allende that W = 1015 n/cm2. ForW = 1015 n/cm2, we obtain 36Cl/35Cl � 4 � 10�8. This issignificant, but far below the values discussed above. Wenote that the meteorite Ningqiang has an exposure age of�40 Myr (Eugster et al., 1988) which would lead to a max-imum expected value of 36Cl/35Cl � 4 � 10�7 which is 1/10of the reported value reported for Ningqiang by Lin et al.(2005). It appears that GCR effects on Allende are insuffi-cient to produce the 36Cl/35Cl levels required. It is furtherof note that the amount of 41Ca produced by secondaryneutrons from GCR exposure is 41Ca/40Ca � 4 � 10�10

for W = 1015 n/cm2, which is much smaller than the valuesreported for Allende. However for W = 1016, this wouldyield 41Ca/40Ca � 0.4 � 10�8 which is comparable to thevalues reported for Allende (Sahijpal et al., 1998).

To get an inferred “initial” 36Cl/35Cl � 10�5 from neu-tron capture would require a net fluence ofW � 3 � 1017 n/cm2 for r35 � 40 barns. If SEP bombard-ment took place on already formed protoplanets, then it isconceivable that this might produce 36Cl by neutron cap-ture. This fluence is only a factor of ten greater than foundin lunar soils for a deposit with an exposure time of�4 � 108 years from GCR (Russ, 1973) but would requirecorrespondingly more intense SEP fluxes. If 36Cl/35Cl at ele-vated levels is present in altered CAIs with low values ofneutron exposure, then it follows that the altered CAIs onany precursor body did not undergo SEP bombardment.If this 36Cl is the result of SEP bombardment and no evi-dence of neutron capture effects are present in the Cl-richalteration phases, then a special environment is requiredthat would ensure the separation of 36Cl from 26Al. If it isa thick target, to avoid large neutron capture effects onany Sm and Gd in the fluid then one possibility is that thetarget was an ice, rich in volatiles (e.g., Cl) and with low dustcontent. This would efficiently produce epithermal neutronsand spallation products of low mass. This then must providethe fluids that altered the CAIs and chondrules. If the targetwere just dust or rocky material enriched in volatiles, thenthe volatile-rich fraction would have to be removed andtransferred to another site to carry out the alteration. Theremaining target would then contain abundant spallationand neutron capture products. We cannot present a self-consistent proposal for the high values of 36Cl/35Cl �10�5

and the absence of 26Al. We suggest that if SEP are respon-sible, then careful monitoring of Sm and Gd isotopes in theCAIs, in the alteration products and associated matrixshould be carried out along with 36Cl investigations. If awater dominated target is involved, further, search for Liand B isotopic anomalies should be looked for in the alter-ation products themselves to test for irradiation products oflow mass elements (oxygen).

In summary, there is reason to consider SEP bombard-ment processes on protoplanets as a possible source ofsome short-lived nuclei. There will be a strong depth depen-dence of the neutron capture rate depending on the depth asample occupied during energetic particle bombardment ofa planet prior to the formation of Allende. There would besome physical separation from the zone where charged

particle interactions dominate to greater depths where neu-trons fluxes are higher. This is quite different from irradia-tion models of energetic particle bombardment of dust inthe solar nebula. Neutron capture in meteorites and withinprotoplanetary surfaces further complicates the interpreta-tion of some isotopic anomalies where very small isotopiceffects have been found in many heavy elements in CAIsusing very high precision techniques.

4.3. Source of Cl and Na in Al3509

The source of the extremely high Cl and Na contents inAl3509 is also a challenging issue. Assuming that 26Al is thedominant heat source at 26Al/27Al � 5 � 10�5, parentbodies of radius greater than a few hundred meters willbe heated. This heating would drive volatiles including hal-ogens and Na, K, Cs outward toward the exterior regionsof any protoplanet. It is expected that hydrous phases,would be the major volatile component. However, there isno evidence for significant aqueous alteration of Al3509.It may therefore necessary to consider that either CO2 orsome other compound might be the mobile component orthat the bulk chemical composition of Al3509 is the fullmetasomatizing agent. The possibility that CO2 is a signif-icant component in the metamorphic processes affectingmeteorites has received relatively little attention, althoughcarbonates (calcite and dolomite) have been found in manycarbonaceous chondrites (Brearley and Jones, 1998 and ref-erences therein). Calcite has been also found surrounding avuggy hibonite mass in the Blue Angel CAI from Murchi-son (Armstrong et al., 1982). However, we note that the sol-ubility of Na and Cl in CO2 is negligible at pressures of upto 70 mPa (Zakirov et al., 2007).

We suggest that Al3509 may possibly represent the bulkfluid with very little water activity, but as indicated above,we cannot infer any further properties of the fluids involved.A study of the phase equilibria for a system with the bulkcomposition of Al3509 (with and without H2O) would be veryimportant in clarifying the matter. It is well known that soda-lite, can be produced under hydrothermal conditions. The is-sue at hand is whether the assemblage of minerals reportedhere is mutually compatible with growth from an aqueousmedium without causing more extensive aqueous alterationto these phases or of the host phases in CAIs. Alternatively,the altered CAIs and Al3509 were later heated, resulting inall water containing phases becoming dehydrated.

5. CONCLUSIONS

We described the mineralogy and petrography of a un-ique Na, Cl, Ca, and Al-rich chondrule Al3509 from the Al-lende meteorite which consists of fine-grained ferroanolivine, ferroan Al-diopside, nepheline, sodalite, and andra-dite, and is crosscut by numerous veins of nepheline, soda-lite, and ferroan Al-diopside.

The Al3509 sodalite shows no evidence of radiogenic26Mg*, establishing an upper limit of the initial 26Al/27Alratio of 3 � 10�6, and suggesting interval for sodalite for-mation of �3 Myr following the crystallization of CVCAIs.


The sodalite grains show large but variable excesses of36S, which, however, do not correlate with 35Cl/34S ratio,possibly indicating local redistribution of radiogenic 36S*

after 36Cl had decayed.Oxygen-isotope compositions of sodalite, ferroan oliv-

ine, and ferroan Al-diopside in Al3509 are similar to thoseof secondary minerals resulting from iron-alkali-halogenmetasomatic alteration of Allende CAIs and chondrulesand in melilite and anorthite of most CAIs in Allende.The oxygen-isotope data suggest that the secondary miner-als precipitated from or equilibrated with a fluid of similaroxygen-isotope composition.

Based on these observations, we infer that the formationof Al3509 and alteration products in CAIs and chondrulesin Allende require a very similar fluid phase, greatly en-riched in volatiles (e.g., Na and Cl). The site for the produc-tion of Na- and Cl-rich fluids responsible for the formationof Al3509 and the alteration of the Allende CAIs and chond-rules must have occurred on a protoplanetary body prior toincorporation into the Allende meteorite; the internal heat-ing of planetesimals by 26Al efficiently could have trans-ferred volatiles to their outer portions and enhanced theformation of volatile-enriched minerals there. We suggestthat the uncorrelated 36S excesses found in Al3509 and thewell correlated 36S excesses reported by other workers maypossibly be the result of the decay of 35Cl resulting from neu-tron capture during SEP bombardment of a fluid rich phaseon the precursor protoplanet which later altered the CAIsand contributed to the debris that now constitutes Allende.

ACKNOWLEDGMENTS

We acknowledge discussions with Lars Borg and MeenakshiWadhwa. The constructive reviews by Makoto Kimura, RogerHewins and Greg Herzog are appreciated. This work was sup-ported by NASA Grants NAG5-10610 and NNX07AI81G (A.N.Krot, P.I.), NAG5-4212 (K. Keil, P.I.), NNG06GG37G (A.J.Brearley, P.I.) and NNH04AB47I (I.D. Hutcheon, P.I.) and bythe Glenn Seaborg Institute. This work was performed under theauspices of the U.S. Department of Energy by Lawrence LivermoreNational Laboratory under Contract DE-AC52-07NA27344. Thisis Hawaii Institute of Geophysics and Planetology PublicationNo. 8210 and School of Ocean and Earth Science and TechnologyPublication No. 8212. G.J. Wasserburg acknowledges support by aNASA Cosmochemistry RTOP to J. Nuth, at GSFC, and by theEpsilon Foundation.

APPENDIX A. SUPPLEMENTARY DATA

Supplementary data associated with this article can befound, in the online version, at doi:10.1016/j.gca.2011.06.004.

REFERENCES

Armstrong J. T., Meeker G. P., Huneke J. C. and Wasserburg G. J.(1982) The Blue Angel. I – the mineralogy and petrogenesis of ahibonite inclusion from the Murchison meteorite. Geochim.

Cosmochim. Acta 46, 575–595.

Bischoff A. and Keil K. (1984) Al-rich objects in ordinarychondrites – related origin of carbonaceous and ordinary

chondrites and their constituents. Geochim. Cosmochim. Acta

48, 693–709.

Brearley A. J. (1997) Disordered biopyriboles, amphibole, and talcin the Allende meteorite; products of nebular or parent bodyaqueous alteration? Science 276, 1103–1105.

Brearley A. J. (1999) Origin of graphitic carbon and pentlandite inmatrix olivines in the Allende meteorite. Science 285, 1380–1382.

Brearley A. J. and Jones R. H. (1998) In Planetary Materials, Rev.

Mineral. Mineral. Soc. Am., Washington, DC. 36, pp. 1–398.Buchanan P. C., Zolensky M. E. and Reid A. M. (1997) Petrology

of Allende dark inclusions. Geochim. Cosmochim. Acta 61,

1733–1743.

Catanzaro E. J., Murphy T. J., Garner E. L. and Shields W. R.(1966) Absolute isotopic abundance ratios and atomic weightsof magnesium. J. Res. Natl. Bur. Stand. 70a, 453–458.

Carlson R. W., Boyet M. and Horan M. (2007) Chondrite barium,neodymium, and samarium isotopic heterogeneity and earlyEarth differentiation. Science 316, 1175–1178.

Choi B.-G., Krot A. N. and Wasson J. T. (2000) Oxygen-isotopesin magnetite and fayalite in CV chondrites Kaba and Mokoia.Meteorit. Planet. Sci. 35, 1239–1249.

Clarke R. S., Jarosewich, Jr., E., Mason B., Nelen J., Gomez M.and Hyde J. R. (1970) The Allende, Mexico, meteorite shower.Smithson. Contrib. Earth Sci. 5, 1–53.

Cosarinsky M., Leshin L. A., MacPherson G. J., Guan Y. andKrot A. N. (2008) Chemical and oxygen isotopic compositionsof accretionary rim and matrix olivine in CV chondrites:constraints on the evolution of nebular dust. Geochim. Cosmo-

chim. Acta 72, 1887–1913.

Duprat J. and Tatischeff V. (2007) Energetic constraints on in situproduction of short lived nuclei in the early solar system.Astrophys. J. 671, l69–l72.

Eugster O., Tera F., Burnett D. S. and Wasserburg G. J. (1970)Neutron capture effects in Gd from the Norton Countymeteorite. Earth Planet. Sci. Lett. 7, 436–440.

Eugster O., Michel Th. and Niedermann S. (1988) Guangnan (L6)and Ningqiang (CV3): Exposure ages and radiogenic ages oftwo unusual chondrites. Meteoritics 23, 25–28.

Eugster O., Herzog G. F., Marti K. and Caffee M. W. (2006)Irradiation records, cosmic-ray exposure ages, and transfertimes of meteorites. In Meteorites and the Early Solar System II

(eds. D. S. Lauretta and H. Y. McSween Jr.). Univ. Arizona

Press, pp. 829–851.

Fireman E. L., Defelice J. and Norton E. (1970) Ages of theAllende meteorite. Geochim. Cosmochim. Acta 34, 873–881.

Fruland R. M., King A. E. and McKay D. S. (1978) Allende darkinclusions. In Proc. Lunar Sci. Conf. 9th. pp. 1305–1329.

Goebel R., Begemann F. and Ott U. (1982) On neutron-inducedand other noble gases in Allende inclusions. Geochim. Cosmo-

chim. Acta 46, 1777–1792.

Goswami J. N., Srinivasan G. and Ulyanov A. A. (1994) Ionmicroprobe studies of Efremovka CAIs: I. Magnesium isotopecomposition. Geochim. Cosmochim. Acta 58, 431–447.

Goswami J. N., Marhas K. K. and Sahijpal S. (2001) Did solarenergetic particles produce the short-lived nuclides present inthe early solar system? Astrophys. J. 549, 1151–1154.

Johnson C. A., Prinz M., Weisberg M. K., Clayton R. N. andMayeda T. K. (1990) Dark inclusions in Allende, Leoville, andVigarano: evidence for nebular oxidation of CV3 constituents.Geochim. Cosmochim. Acta 54, 819–831.

Hashimoto A. and Grossman L. (1987) Alteration of Al-richinclusions inside amoeboid olivine aggregates in the Allendemeteorite. Geochim. Cosmochim. Acta 51, 1685–1704.

Hidaka H., Ebihara M. and Yoneda S. (2000) Isotopic study ofneutron capture effects on Sm and Gd in chondrites. Earth

Planet. Sci. Lett. 180, 29–37.




Hidaka H. and Yoneda S. (2009) Isotopic evidence of non-thermalized neutron irradiation in solar–gas-rich meteorites:possibility of the interaction with solar neutrons and activityfrom the early Sun. Earth Planet. Sci. Lett. 285, 173–178.

Hsu W., Guan Y., Leshin L. A., Ushikubo T. and Wasserburg G.J. (2006) A late episode of irradiation in the early solar system:evidence from extinct 36Cl and 26Al in meteorites. Astrophys. J.

640, 525–529.

Ikeda Y. and Kimura M. (1995) Anhydrous alteration of Allendechondrules in the solar nebula I: description and alteration ofchondrules with known oxygen-isotopic compositions. Proc.

NIPR Symp. Antarct. Meteorites 8, 97–122.

Ishii H. A., Krot A. N., Bradley J. P., Keil K., Nagashima K.,Teslich N., Jacobsen B. and Yin Q.-Z. (2010) Discovery,mineral paragenesis and origin of wadalite in a meteorite. Amer.

Mineral. 95, 440–448.

Jacobsen B., Yin Q.-Z., Moynier F., Amelin Y., Krot A. N.,Nagashima K., Hutcheon I. D. and Palme H. (2008) 26Al–26Mgand 207Pb–206Pb systematics of Allende CAIs: canonical solarinitial 26Al/27Al ratio reinstated. Earth Planet. Sci. Lett. 272,

353–364.

Jacobsen B., Matzel J., Hutcheon I. D., Ramon E., Krot A. N.,Ishii H., Nagashima K. and Yin Q.-Z. (2009) The 36Cl–36Ssystematic of CAI wadalite from the Allende meteorite:evidence for possible early incorporation of live 36Cl into CAIs(abstract). Lunar Planet. Sci. 40, #2553.

Jogo K., Nakamura T., Noguchi T. and Zolotov M. Yu. (2009)Fayalite in the Vigarano CV3 carbonaceous chondrite: Occur-rences, formation age and conditions. Earth Planet. Sci. Lett.

287, 320–328.

Kennedy A. K. and Hutcheon I. D. (1994) Chemical and isotopicconstraints on the formation and crystallization of SA-1, abasaltic Allende plagioclase-olivine inclusion. Meteoritics 27,

539–554.

Kimura M. and Ikeda Y. (1995) Anhydrous alteration of Allendechondrules in the solar nebula; II, Alkali-Ca exchange reactionsand formation of nepheline, sodalite and Ca-rich phases inchondrules. Proc. NIPR Symp. Antarct. Meteorites 8, 123–138.

Kimura M. and Ikeda Y. (1998) Hydrous and anhydrous altera-tions of chondrules in Kaba and Mokoia CV chondrites.Meteorit. Planet. Sci. 33, 1139–1146.

Kojima T. and Tomeoka K. (1996) Indicators of aqueousalteration and thermal metamorphism on the CV3 parentbody: microtextures of a dark inclusion from Allende. Geochim.

Cosmochim. Acta 60, 2651–2666.

Krot A. N. and Rubin A. E. (1994) Glass-rich chondrules inordinary chondrites. Meteoritics 29, 697–707.

Krot A. N., Scott E. R. D. and Zolensky M. E. (1995) Mineralogicand chemical variations among CV3 chondrites and theircomponents: nebular and asteroidal processing. Meteoiritics 30,

748–775.

Krot A. N., Petaev M. I., Scott E. R. D., Choi B.-G., Zolensky M.E. and Keil K. (1998a) Progressive alteration in CV3 chon-drites: more evidence for asteroidal alteration. Meteorit. Planet.

Sci. 33, 1065–1085.

Krot A. N., Zolensky M. E., Keil K., Scott E. R. D. andNakamura K. (1998b) Secondary Ca–Fe-rich minerals in theBali-like and Allende-like oxidized CV3 chondrites and Allendedark inclusions. Meteorit. Planet. Sci. 33, 623–645.

Krot A. N., Hiyagon H., Petaev M. I. and Meibom A. (2000)Oxygen isotopic compositions of secondary Ca–Fe-rich silicatesfrom the Allende dark inclusions: evidence against high-temperature formation (abstract). Lunar Planet. Sci. 31, #1463.

Krot A. N., Petaev M. I., Meibom A. and Keil K. (2001) In situ

growth of Ca-rich rims around Allende dark inclusions.Geochem. Int. 36, 351–368.

Krot A. N., Hutcheon I. D. and Keil K. (2002) Anorthite-richchondrules in the reduced CV chondrites: evidence for complexformation history and genetic links between CAIs and ferro-magnesian chondrules. Meteorit. Planet. Sci. 37, 155–182.

Krot A. N., Yurimoto H., Hutcheon I. D., Libourel G., Chaussi-don M., Petaev M. I., MacPherson G. J., Paque-Heather J. andWark D. (2007) Anorthite-rich, igneous (Type C) Ca, Al-richinclusions from the CV carbonaceous chondrite Allende:evidence for multistage formation history. Geochim. Cosmo-

chim. Acta 71, 4342–4364.

Krot A. N., Nagashima K., Hutcheon I. D., Ishii H. A., JacobsenB., Yin Q.-Z., Davis A. M. and Simon S. B. (2010) Mineralogy,petrography, oxygen and magnesium isotopic compositions andformation age of grossular-bearing assemblages in the AllendeCAIs (abstract). Lunar Planet. Sci. 41, #1441.

Kurat G., Palme H., Brandstatter F. and Huth J. (1989) Allendexenolith AF: undisturbed record of condensation and aggrega-tion of matter in the solar nebula. Zeitschr. Naturforsch. 44a,

988–1004.

Leya I., Halliday A. N. and Wieler R. (2003) Predictableconsequences of nucleosynthesis by spallation reactions on theearly solar system. Astrophys. J. 594, 605–616.

Lin Y., Guan Y., Leshin L. A., Ouyang Z. and Wang D. (2005)Short-lived chlorine-36 in a Ca- and Al-rich inclusion from theNingqiang carbonaceous chondrite. Proc. Natl. Acad. Sci. 102,

1306–1311.

Mason B. and Martin P. M. (1977) Geochemical differences amongcomponents of Allende meteorite. Smithson. Contrib. Earth Sci.

19, 84–95.

Mason B. and Taylor S. R. (1982) Inclusions in the Allendemeteorite. Smithson. Contrib. Earth Sci. 25, 1–30.

Matzel J. E. P., Jacobsen B., Hutcheon I. D., Krot A. N.,Nagashima K., Yin Q.-Z., Ramon E. C., Weber P. K. andWasserburg G. J. (2010) Distribution and origin of 36Cl inAllende CAIs (abstract). Lunar Planet Sci. 41, #2631.

Nakashima D., Ott U., Hoppe P. and El Goresy A. (2008) Searchfor extinct 36Cl: Vigarano CAIs, the Pink Angel from Allende,and a Ningqiang chondrule. Geochim. Cosmochim. Acta 72,

6141–6153.

Nakashima D., Ott U. and Hoppe P. (2010) Search for extinctchlorine-36: a halite grain from the Zag meteorite (abstract).Meteorit. Planet. Sci. 45, #5289.

Nehru C. E., Weisberg M. K., Ebel D. S., Boesenberg J. S. andConnolly, Jr., H. C. (2008) Origin of Na-, Al-, glass-richchondrules in H, L and LL chondrites (abstract). Lunar Planet.

Sci. 39, #1697.

Nishiizumi K., Elmore D., Kubik P. W., Bonani G., Suter M.,Wolfli W. and Arnold J. R. (1986) Age of Antarctic meteoritesand ice II. Lunar Planet. Sci. 17, 18–20.

Pouchou J. L. and Pichoir F. (1984) A new model for quantitativeX-ray microanalysis. Part I: application to the analysis ofhomogeneous samples. Rcch. Aerosp. 3, 13–38.

Pravdivtseva O. V., Krot A. N., Hohenberg C. M., Meshik A. M.,Weisberg M. K. and Keil K. (2003) The I-Xe record ofalteration in the Allende CV chondrite. Geochim. Cosmochim.

Acta 67, 5011–5026.

Reynolds J. H. (1963) Xenology. J. Geophys. Res. 68, 2939–2956.

Russ, III, G. P. (1973) Apollo 16 neutron stratigraphy. Earth


Russ, III, G. P., Burnett D. S., Lingenfelter R. E. and WasserburgG. J. (1971) Neutron capture on 149Sm in lunar samples. Earth


Russell S. S., Srinivasan G., Huss G. R., Wasserburg G. J. andMacPherson G. J. (1996) Evidence for widespread 26Al in thesolar nebula and constraints for nebula time scales. Science 273,

757–762.


Sahijpal S., Goswami J. N., Davis A. M., Lewis R. S. andGrossman L. (1998) A stellar origin for the short-lived nuclidesin the early solar system. Nature 391, 559–561.

Srinivasan G., Huss G. R. and Wasserburg G. J. (1999) Apetrographic, chemical, and isotopic study of calcium–alumi-num-rich inclusions and aluminum-rich chondrules from theAxtell (CV3) chondrite. Meteorit. Planet. Sci. 35, 1333–1354.

Turner G., Crowther S. A., Burgess R., Wasserburg G. J., Kelley S.P. and Gilmour J. D. (2009) Pink Angel: argon and xenondiffusion, I-Xe chronology and the 36Cl problem (abstract).Meteorit. Planet. Sci. 44, #5255.

Ushikubo T., Guan Y., Hiyagon H., Sugiura N. and Leshin L. A.(2007) 36Cl, 26Al, and O isotopes in an Allende type B2 CAI:implications for multiple secondary alteration events in theearly solar system. Meteorit. Planet. Sci. 42, 1267–1279.

Villa I. M., Huneke J. C., Papanastassiou D. A. and WasserburgG. J. (1981) The Allende Pink Angel: chronological constraintsfrom Xe, Ar, and Mg (abstract). Lunar Planet. Sci. 12, 1115–

1117.

Wark D. A. and Lovering J. F. (1977) Marker events in the earlysolar system: evidence from rims on Ca–Al-rich inclusions incarbonaceous chondrites. Proc. Lunar Sci. Conf. 8, 95–112.

Wasserburg G. J. (1985) Short lived nuclei in the early solar system.In Protostars and Planets II. Univ. Arizona Press, pp. 703–737.

Wasserburg G. J. and Huneke J. C. (1979) I-Xe dating of I-bearingphases in Allende (abstract). Lunar Planet. Sci. 10, 1307–1309.

Wasserburg G. J., Busso M., Galino R. and Nollett K. M. (2006)Short-lived nuclei in the early Solar System: possible AGBsources. Nucl. Phys. A 777, 5–69.

Weisberg M. K. and Prinz M. (1998) Fayalitic olivine in CV3chondrite matrix and dark inclusions: a nebular origin. Mete-

orit. Planet. Sci. 33, 1087–1111.

Whitby J., Burgess R., Turner G., Gilmour J. and Bridges J. (2000)Extinct 129I in halite from a primitive meteorite: evidence forevaporite formation in the early solar system. Science 288,

1819–1821.

Zakirov I. V., Sretenskaja N. G., Aranovich L. Y. and Volchenk-ova V. A. (2007) Solubility of NaCl in CO2 at high pressure andtemperature: first experimental measurements. Geochim. Cos-

mochim. Acta 71, 4251–4255.

Zolensky M. E., Bodnar R. J., Gibson E. K., Gounelle M., NyquistL. E., Reese Y., Shih C.-Y. and Wiesmann H. (1999) Asteroidalwater within fluid inclusion-bearing halite in an H5 chondrite.Science 285, 1377–1379.

Associate editor: Gregory Herzog

Extremely Na- and Cl-rich chondrule from the CV3 ...starplan.dk/sites/starplan.dk/files/wasserburg, et al.pdf · Extremely Na- and Cl-rich chondrule from the CV3 carbonaceous chondrite

Documents