Special Reports

Isotopic Compositions of Cometary Matter Returned by Stardust

See allHide authors and affiliations

Science  15 Dec 2006:
Vol. 314, Issue 5806, pp. 1724-1728
DOI: 10.1126/science.1135992

Abstract

Hydrogen, carbon, nitrogen, and oxygen isotopic compositions are heterogeneous among comet 81P/Wild 2 particle fragments; however, extreme isotopic anomalies are rare, indicating that the comet is not a pristine aggregate of presolar materials. Nonterrestrial nitrogen and neon isotope ratios suggest that indigenous organic matter and highly volatile materials were successfully collected. Except for a single 17O-enriched circumstellar stardust grain, silicate and oxide minerals have oxygen isotopic compositions consistent with solar system origin. One refractory grain is 16O-enriched, like refractory inclusions in meteorites, suggesting that Wild 2 contains material formed at high temperature in the inner solar system and transported to the Kuiper belt before comet accretion.

The isotopic compositions of primitive solar system materials record evidence of chemical and physical processes involved in the formation of planetary bodies ∼4.6 billion years ago and, in some cases, provide a link to materials and processes in the molecular cloud that predated our solar system. The vast majority of isotopic analyses of extraterrestrial materials have been performed on chondritic (undifferentiated) meteorites, samples of asteroids that likely accreted at 2 to 4 astronomical units (AU) within the first few million years of solar system history. Comets formed in much colder regions of the protoplanetary disk and are widely considered to consist of more primitive matter than even the most unequilibrated meteorites.

Analyses of isotope compositions of comets are rare. Measurements of D/H, 13C/12C, 15N/14N, or 18O/16O have been made for a few abundant molecules in gases of several comet comae by ground-based spectroscopy (13) and of comet P/Halley by mass spectrometers on the Giotto spacecraft (4). Direct measurements of isotope compositions in the dust fraction of comets are limited to low-precision data from the Halley flyby (5). Isotopic measurements of stratosphere-collected interplanetary dust particles (IDPs) demonstrate the highly primitive nature of many anhydrous IDPs [e.g., (6, 7)]; however, a cometary origin for specific individual particles cannot be ascertained. Here, we report laboratory analyses of the light “stable” isotopes of H, C, N, O, and Ne in individual grains, particle fragments, crater debris, and/or trapped volatile materials collected from comet 81P/Wild 2 and returned to Earth by the NASA Discovery Mission, Stardust.

The goals of the Isotope Preliminary Examination Team analyses are to provide first-order answers to questions relating to the provenance of Wild 2 dust: (i) Does the comet consist of a mechanical agglomeration of essentially unprocessed, or perhaps only thermally annealed, presolar materials? (ii) Do comets provide a well-preserved reservoir of circumstellar dust grains (8) with distinct nucleosynthetic histories (i.e., stardust)? (iii) Can isotopic signatures establish whether extraterrestrial organic materials are present above contamination levels? (iv) What are the relations to known isotope reservoirs in meteoritic samples and in IDPs? (v) What are the implications for mixing and thermal processing in the early solar system? The preliminary examination has focused on the abundant light elements H, C, N, and O and on light noble gases because they have characteristic isotopic signatures that vary widely in solar system materials [e.g., (9)] and compositions that can often be linked to distinct astrophysical environments and/or processes.

Hydrogen isotopic compositions in bulk fragments of five Wild 2 particles (see supporting online material) range from values similar to those of terrestrial and chondritic particles, up to moderate D/H enhancements of about three times the D/H in mean ocean water (Fig. 1, table S1). Variations in δD within a particle are also observed with some “hotspots” having a δD up to 2200 ± 900 per mil (‰) (10). All the observed D enrichments are associated with carbon, although only one of the samples (T17) is dominated by carbonaceous material. Although overlapping the range of values observed for IDPs (11, 12) and water in(Oort cloud) comets (4), the maximum D/H is well below that characteristic of a minor component of organic matter in anhydrous IDPs, especially of the low-density “cluster” type thought to be derived from comets (13). The maximum D/H values of Wild 2 grains are also far less than those measured in small hotspots of insoluble organic matter separated from carbonaceous chondrites (14) and in cometary HCN ice (15), which are both thought to have close affinities to organic material produced at very low temperatures in molecular cloud environments. It is not known if this relative paucity of highly D-enriched matter signifies an intrinsic difference between Wild 2 samples and macromolecular materials in IDPs and carbonaceous chondrites. It is also possible that D/H signatures were modified during impact. Because of the association with carbon and the lack of evidence for any hydrated minerals in Wild 2 particles, it is unlikely that D/H values can be ascribed to Wild 2 water.

Fig. 1.

Hydrogen isotopic compositions in bulk fragments (solid circles) of five Wild 2 particles and in micrometer-sized subareas of one particle (open circles) compared to SMOW and to ranges of laboratory measurements of D/H in IDPs and in insoluble organic matter (IOM) from chondritic meteorites. Also shown are an estimate for protosolar H2 and ranges of D/H measured remotely for specific gaseous molecules from comets and for molecular clouds. Error bars are 1σ.

Some cometary volatiles appear to have been captured. Helium, neon, and argon abundances, as well as 20Ne/22Ne ratios, were analyzed in two bulbous sections of an impact track that contains fragments of fine-grained impactor material mixed with melted aerogel along its periphery. Noble gases in a control sample (unexposed aerogel taken from the rear portion of a collector cell) are consistent with system blanks, indicating that flight aerogel does not contain a detectable background of noble gases. In contrast, aerogel fragments containing the impact track show excess He and Ne above blank levels by factors of 3 to 4 (table S2), suggesting that the very rapid time scales (<μs) for melting of aerogel during deceleration of comet particles helped trap indigenous cometary volatiles. This inference is supported by measured 20Ne/22Ne ratios, which are significantly higher than that in air. In a 22Ne/20Ne versus 4He/20Ne diagram (fig. S1), the data do not lie on a pure mixing trajectory between air and the solar composition but instead point toward a nonatmospheric end-member with a 4He/20Ne ratio lower than solar, consistent with expectations for comets due to preferential trapping of heavier noble gases in ice accreted at ∼30 K (16).

Carbon and nitrogen isotope analyses were made by isotopic mapping with NanoSIMS instruments, measuring atomic ions of C and isotopologs of the intense CN beam, respectively. Microtomed particles extracted from aerogel, as well as Al-foil crater debris, were mapped with a spatial resolution of ∼100 nm to search for isotopic anomalies that could identify circumstellar dust grains like the C- and N-rich phases (SiC, Si3N4, graphite) found in primitive meteorites (17). No circumstellar dust grains were definitively identified despite an intensive search comprising more than 20 slices and fragments from a dozen particles and ∼5700 μm2 of debris in and around six small (<2 μm) and four large (>50 μm) craters on Al-foil collectors. This result contrasts sharply with inferences of a population of nearly pure 12Cgrains in comet Halley (5). One ∼150-nm region from the bottom of a large Al-foil crater has an isotopic composition (δ13C=59 ±61‰; δ15N = –518 ± 60‰) that falls in a range consistent with “mainstream” presolar SiC grains (17); however, the grain was sputtered away before a mineral identification could be made. Another hotspot was enriched in 13C(δ13C=964±219‰) and depleted in 15N(δ15N = –415 ± 94‰), typical of mainstream SiC; however, the ion-emission area was not as localized as would be expected for a very small circumstellar grain and, moreover, showed a high abundance of nitrogen (inferred C/N ∼3). This grain disappeared rapidly during analysis, as did a second spot with a low inferred C/N ratio ≈ 6 and low δ15N ≈–350‰. This behavior is consistent with sputtering of labile organic material. Notably, this isotopically light nitrogen component is rarely seen in IDPs (7, 18), but within ∼10% uncertainties, is consistent with estimates of the nitrogen isotopic composition in HCN gas from comet Hale Bopp (19) and of the solar composition based on analyses of the jovian atmosphere (20) and of solar wind implanted in lunar grains (21).

On a micrometer scale, all samples are homogeneous in both C and N isotope compositions and show no correlation between δ13C and δ15N (Fig. 2). Carbon isotopes fall in a particularly restricted range with most “bulk” δ13C values falling between –20 and –50‰. This range is somewhat higher than recent estimates (δ13C = –105 ± 20‰) of the solar carbon isotope composition (22), but is compatible with carbon isotopes in other primitive solar system matter including IDPs (6) and most organic matter in chondrites (23).

Fig. 2.

Correlated carbon and nitrogen isotopic compositions of individual cometary grains compared with literature data (gray circles) for stratosphere-collected IDPs (7). Shown are average values for cometary grains prepared as microtome sections (open squares), for fragments extracted from aerogel and pressed into Au foil (open circles), and for residue in and surrounding craters in the Al foil collector (solid circles). Also shown are selected hotspots (solid squares) from the crushed samples. Error bars are 2σ.

A wider range is observed for “bulk” nitrogen with many samples clustering near 0‰ (atmospheric) but others showing modest δ15N enrichments of + 100 up to ∼500‰, similar to the range observed for anhydrous IDPs (7). On a submicrometer scale, more-extreme values are found with amaximum δ15N of 1300 ± 400‰, similar to the highest values found in refractory organic matter in IDPs (7) and meteorites (14). Unlike IDPs, however, the Wild 2 samples display both low and high δ15N, indicating an unequilibrated mixture of a low-δ15N (perhaps icy) component and a more refractory (high C/N) organic material with high δ15N and isotopically “normal” carbon. The fact that most IDPs are characterized by 15N excesses may reflect the instability of the more labile, low-δ 15N component during atmospheric-entry heating. Thus, the Wild 2 samples could represent a different type of organic material than that available for study through the stratospheric IDPs, although it is puzzling that the Wild 2 particles analyzed thus far appear to be deficient in total organic matter compared to most IDPs. This observation contradicts expectations of comets as being very rich in organic matter and may indicate that much of the Wild 2 organic matter did not survive the capture process as discrete phases still closely associated with silicate minerals.

Oxygen is the most abundant element in rocky planets, asteroids, and comets and exhibits distinctive isotopic compositions that are essentially unique to each class of planetary materials from the inner solar system (24). Although the isotopic variations that occur on planetary scales are relatively subtle (less than a few ‰), individual components of unequilibrated meteorites have oxygen isotope compositions varying by 50‰ or more in relative abundances of 16O[e.g., (9)]. These isotopic variations are fairly systematic (25) and, together with planetary-scale variations, imbue oxygen with a unique diagnostic capability to indicate sample provenance (24). Orders-of-magnitude larger isotopic variations are observed in presolar oxide and silicate dust grains that condensed in the outflows of evolved, mass-losing stars and inherited specific isotopic compositions due to local nucleosynthetic processes (26). Few definitive data exist regarding oxygen isotope compositions in cometary materials. In situ measurements of water ice from comet Halley, made by the Giotto mission, yield δ18O=12 ± 75‰, but no measurement of δ17O is available (4).

Oxygen isotope measurements of Wild 2 samples were made by mapping techniques to search for presolar grains and by focused probe SIMS on individual particle fragments to relate the comet samples to known classes of meteoritic materials. Of some 2 dozen particle fragments prepared as microtomed (<200 nm thick) sections from aerogel keystones, no candidate presolar grains were identified on the basis of extreme oxygen isotope anomalies (supporting online material). For several samples, the measurement of oxygen isotopes was compromised by mixing with melted aerogel, which could not be resolved even with the <100-nm spatial resolution of the NanoSIMS. Therefore, residues from impact craters in the Al-foil targets were also examined by high-resolution isotope mapping.

Thirty-seven small craters, between 320 nm and 1.5 μm in diameter, and four large craters (59, 72, 140, and >200 μm in diameter) were mapped, resulting in the identification of ∼104 O-rich subareas. Only one presolar grain was found (Fig. 3). The ∼250-nm grain is highly enriched in 17O and slightly depleted in 18O compared to solar system samples and has a composition of 17O/16O = (1.01 ± 0.20) × 10–3 and 18O/16O = (1.77 ± 0.24) × 10–3. This isotopic composition is typical for presolar oxide (and silicate) grains belonging to “group 1,” thought to originate in red giant or asymptotic giant-branch stars (26). Unfortunately, the mineralogy of the grain could not be determined, although it is likely that it was a relatively refractory oxide or silicate.

Fig. 3.

Presolar grain discovered in residue of crater C2086W1. (A) Backscattered electron image of a crater which punctured the Al foil. (B) Electron image of projectile residue in crater lip. (C) False-color isotope map of δ17O measured by high-resolution NanoSIMS. (D): Oxygen three-isotope plot of presolar grain (open square) compared to literature data for presolar oxide grains separated from meteorites. On this scale, all materials formed in the solar system plot at the intersection of the two dashed lines. Error bars are 1σ.

High-precision oxygen isotope measurements were made in 5- to 10-μm spots of individual “terminal grains” separated from aerogel tracks and pressed into clean Au foil. Fragments from one enstatite-rich grain (T69) and one forsterite-rich grain (T22) have similar oxygen isotope compositions that plot slightly below the terrestrial mass-dependent fractionation (TF) line and to the low δ18O side of the 16O-mixing line that characterizes refractory components (calcium- and aluminum-rich inclusions, CAIs) in chondrites (Fig. 4; table S2). Although unlikely, we cannot exclude the possibility that the deviation of the T69 and T22 data from the CAI mixing line is due to minor contamination with aerogel (supporting online material). The oxygen isotopic compositions of these Mg-rich silicates from Wild 2 are compatible with those of most mafic silicate minerals from carbonaceous chondrite chondrules.

Fig. 4.

Oxygen isotopic compositions of individual cometary grains. The terrestrial mass-dependent fractionation line (TF) and the CAI-mixing line are shown for reference. Error bars are 2σ.

Oxygen isotopes were also measured in a polymineralic refractory grain (T25, “Inti”) by means of both NanoSIMS mapping and higher-precision IMS 1270 spot measurements (supporting online material). The sample consists of a fine-grained mix of spinel, Al-rich diopside, melilite, and anorthite, with a minor component of perovskite (27). Although the NanoSIMS analyses resolve individual phases, it was not possible to avoid sampling multiple mineral phases during the IMS 1270 analyses (5-μm-diameter spot). The consistency of the data (Fig. 4) demonstrates that the sample is isotopically homogeneous with an 16O-rich composition compared to most planetary materials (δ18O ≈ δ17O ≈–40‰). Notably, this composition is virtually identical to that of a large population of CAIs and refractory oxide grains in chondrites (9). Similarly 16O-enriched compositions have been observed for rare refractory IDPs collected in the stratosphere (28), but the relation between these particles and meteoritic CAIs has not been thoroughly investigated.

As an ensemble, the isotopic compositions of the light elements—H, C, N, O, and Ne—demonstrate that the dust of comet Wild 2 is an unequilibrated aggregate of material from different sources. The H, C, and N isotope compositions indicate the presence of several minor components that are isotopically fractionated to a large extent, probably reflecting chemical reactions at very low temperatures like those characteristic of molecular cloud or, possibly, Edgeworth-Kuiper belt environments. This is consistent with a general view of Jupiter family comets as having accreted from cold materials at the edge of the solar nebula.

Two observations, however, run counter to expectations: (i) The abundance of presolar grains appears to be low compared with that of primitive meteorites and IDPs, and (ii) the comet contains high-temperature silicate and oxide minerals with oxygen isotopic compositions essentially identical to those of analogous minerals in carbonaceous chondrites. The first observation could perhaps be explained as a preservation effect, i.e., loss of presolar materials during impact, yet the one presolar grain firmly identified was found in the residue of the largest impact crater so far investigated (the grain actually punctured the Al foil). If similar grains were abundant, they should have been seen in or around other craters and in the aerogel as well. However, if most of the presolar material in Wild 2 consists of interstellar amorphous silicates rather than circumstellar mineral grains, then it is possible that these would not be recognized either because of too much mixing with aerogel or, in the case of residue in Al-foil craters, because they might not possess sufficiently anomalous oxygen isotopic compositions to be identified as nonsolar. On the other hand, the crystalline silicate and oxide minerals for which we have oxygen isotope data could not have formed by annealing (devitrification) of presolar amorphous silicates in the Kuiper belt. Not only is such an origin incompatible with the chemistry and mineralogy of these grains, but, because these grains differ markedly in their relative 16O abundances, they also could not have formed from a single isotopic reservoir. The similarity in O-isotope composition between Wild 2 grains and materials from carbonaceous chondrites is pronounced. Identifying precisely what regions of the inner solar system may be sampled by this comet is clearly central to understanding the scale of radial mixing outward in the solar protoplanetary disk.

Most interstellar silicate dust is thought to be amorphous (29), and with recent recognition of the relatively large fraction of crystalline silicates in comets and in protostellar disks based on infrared spectral data [e.g., (30)], astronomers have postulated mixing of dust from inner warm regions, where ambient conditions are above glass transition temperatures, outward to Kuiper belt regions [e.g., (31, 32)]. However, other observations suggest production of crystalline silicates (specifically forsterite) at large (>10 AU) stellar distances (33). That theCAI-likeparticle(Inti) has the same intrinsic oxygen isotopic properties as CAIs from all major classes of primitive meteorites argues strongly for an origin in the same source region as meteoritic CAIs (9). One possibility is that this region could be at the inner truncation of the accretion disk (<0.1 AU) where very high temperatures exist. Here, winds associated with bipolar outflows, driven by interactions of the young Sun with the accretion disk, can carry small refractory particles to large heliocentric distances where they can accrete together with cold, icy materials (34, 35). Other models (36) invoke turbulent transport of hot inner nebula (silicate) dust out to zones of comet accretion, which could account for the carbonaceous chondrite-like isotopic compositions of Wild 2 olivine and pyroxene grains. We conclude that the coupled oxygen isotopic and mineralogic data are best understood as indicating that a large fraction of dust in comet Wild 2 is derived from chemically and thermally processed precursors from the inner solar system, consistent with predictions of the X-wind and other models (3436) for the protosolar disk.

Supporting Online Material

www.sciencemag.org/cgi/content/full/314/5806/1724/DC1

Methods

Fig. S1 to S5

Tables S1 and S6

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article