INTERPLANETARY DUST PARTICLES AS SAMPLES OF ICY ASTEROIDS

Observations of more than 100,000 main belt asteroids (MBAs) in visible wavelengths show that ∼60% of the mass of the asteroid belt belongs to the so-called C-complex, which comprises six subclasses (B, C, Cb, Cg, Cgh, Ch; DeMeo et al. 2009; DeMeo & Carry 2013). While one-third of all C-complex asteroids (Ch- and Cgh-types; Rivkin 2012) possess a well-established meteoritic analog (CM chondrites; see Figure 4; e.g., Cloutis et al. 2011), the bodies in the remaining subclasses (B-, C-, Cb- and Cg-type), including Ceres, appear unsampled by our meteorite collections, similar to P- and D-type asteroids (see A.1 and A.2). Importantly, whereas metamorphosed CI/CM chondrites have been proposed in the past as analogs of B-, C-, Cb- and Cg-type surfaces (Hiroi et al. 1993), such a possibility is presently untenable for the majority of these asteroids for three reasons, namely the paucity of these meteorites among falls (∼0.2% of meteorite falls) compared to the abundance of B-, C-, Cb- and Cg-types, the difference in density between these meteorites and these asteroids, and the difference in spectral properties (in the mid-infrared) between the two groups (see A.1 for discussion). Overall, it appears that the surface material of ∼66% of the MBAs masses (B-, C-, Cb-, and Cg-types—BCG-types in the following—represent ∼53% and P- and D-types represent ∼13% of the main belt's mass; DeMeo & Carry 2013) are absent from our meteorite collections.

A plausible explanation to this paradox could be the "comet-like" nature of these objects (see A.3). Indeed, the few existing density measurements (∼0.8 < p < ∼1.5 g cm⁻³ for D < 200 km bodies; Carry 2012; Marchis et al. 2012) indicate that these bodies cannot be made of silicates only, and must comprise a significant fraction of ice(s). This "comet-like" nature applies also to the dwarf planet Ceres as plumes of water vapour have unambiguously been detected at this object (Küppers et al. 2014; Ceres'mass ∼35% of the main belt mass). It is thus not a surprise that several so-called main-belt comets were discovered among BCG-, P- and D-type objects (e.g., Hsieh & Jewitt 2006; Jewitt 2012; Hui & Jewitt 2015). Additional support to the "comet-like" nature of these objects is provided by spectroscopy in the 3 μm region, which indicates that water ice may not only be present in the interior of these bodies but also on their surfaces, as in the case of the Themis family (Campins et al. 2010; Rivkin & Emery 2010; Licandro et al. 2011; Hargrove et al. 2012, 2015; Takir & Emery 2012).

Because a large fraction of MBAs appears unsampled by our meteorite collections (at least their surface material), it seems logical, as a next step, to test a link between these asteroids and the other significant source of extraterrestrial materials, namely interplanetary dust particles (IDPs). IDPs differ from meteorites in being smaller (<2 mm), more plentiful (they contribute most of the extraterrestrial material that comes to the present-day Earth) and different in texture and composition (Bradley 2005). In particular, some classes of IDPs appear to be the most primitive material in the solar system and at present provide our best source of information on the nature and evolution of the particles in the preplanetary solar nebula (Bradley 1999, pp. 485–503). While both comets and asteroids are proposed as the most important contributors to the influx of IDPs, the relative contributions of these two populations to the IDP influx as well as the nature of the parent bodies of the three main classes of IDPs (chondritic porous pyroxene-rich or olivine-rich IDPs, and phyllosilicate-rich IDPs; Bradley 2005) remain elusive. Presently, existing spectral datasets open the possibility that hydrous IDPs originate from C-complex asteroids (Bradley et al. 1996) and that anhydrous chondritic porous (pyroxene-rich and/or olivine-rich) IDPs originate from P- and D-type asteroids (although this proposed link relies on the visible range alone; Bradley et al. 1996) and/or comets (Sandford & Walker 1985; Campins & Ryan 1989; Sandford & Bradley 1989; Wooden 2002, 2008; Brunetto et al. 2011; Schulz et al. 2015). Modeling work based on observations of solar system dust bands suggests that either asteroids (e.g., Dermott et al. 2002 and references therein) or comets (e.g., Nesvorný et al. 2010) are the dominant source of zodiacal dust and thus of IDPs, leaving both possibilities open for further investigation.

The structure of this paper is as follows. In Section 2, we compare the spectral properties of BCG-, P- and D-type asteroids on the one hand, and those of comets on the other hand, with those of IDPs over a very large wavelength range (0.4–13 μm). In Section 3, we present the implications of our spectral analysis regarding the origin of IDPs. Finally, we list in Section 4 the remaining issues regarding both the origin of IDPs and the asteroid–comet-IDP link that require critical new data.

To explore the composition of the enigmatic icy asteroids (BCG-, P- and D-type asteroids) in more detail as well as a possible link with IDPs, we first compared their spectra in both the visible and mid-infrared ranges with those of IDPs (as of today, analog data in the 0.8–2.5 μm range unfortunately does not exist for IDPs). Spectra in these two wavelength ranges were available from previously published studies (see the figure captions for the complete list of references). In both ranges, it appears that the spectral properties of BCG-, P- and D-type asteroids are compatible with those of chondritic porous (anhydrous) IDPs (Figure 1). The mid-infrared spectral range, which is known to be significantly more powerful for the determination of the composition of silicate assemblages than the 0.4–0.8 μm range alone (Vernazza et al. 2011, 2012, 2013), further allows us to pinpoint the likely parent bodies of the two classes of chondritic porous IDPs: BCG-type asteroids have spectral properties compatible with pyroxene-rich IDPs, while P-types, D-types and comets have spectral properties compatible with a mixture of pyroxene-rich and olivine-rich IDPs (Figure 1). Note that P-types and some comets (see Figures 6 and 8 as well as A.7) seem to fill the gap that separates the pyroxene-rich BCG-types from the more olivine-rich D-types as well as most comets.

Zoom In Zoom Out Reset image size

As a next step, we verified that the proposed link between BCG-, P- and D-type asteroids and chondritic porous IDPs is still valid in the near-infrared spectral range (0.8–2.5 μm) for which IDP spectra are currently missing. To test this link, we applied a radiative transfer model based on the Mie and Hapke theory (see A.7 for a detailed description of the model) to a subset of asteroid spectra representative of the spectral diversity seen among BCG-, P- and D-type asteroids (DeMeo et al. 2009; we actually applied the model to both the 0.45–2.45 μm and 8–13 μm ranges in order to cover the entire existing spectral range). The near-infrared asteroid spectra were collected with the SpeX instrument on the NASA Infrared Telescope Facility (IRTF) on Mauna Kea, Hawaii (see A.4 and A.5).

As end-member, we used all the major and most minor minerals found in chondritic porous IDPs (Bradley 2005; crystalline olivine and pyroxene, silicate glasses, amorphous carbon, sulfides such as pyrrothite and FeNi metal). To reproduce the spectral slope of P- and D-type asteroids we used the space-weathering model of Brunetto et al. (2006). Our results confirm that the spectral properties of BCG, P- and D-type asteroids are compatible in both investigated spectral ranges with their surfaces being covered by fine-grained silicates (grain size in the ∼0.2–1 μm range as observed in IDPs) that have chondritic porous IDP-like compositions (see Figures 2 and 8).

Zoom In Zoom Out Reset image size

If we follow the reasonable assumption proposed in the literature that hydrous IDPs originate from water-rich (or ice-rich) asteroids (e.g., Sandford & Bradley 1989; Dermott et al. 2002, pp. 423–442; Bradley 2005), then our study implies that asteroids (both their surfaces and their interiors) are fully representative of the compositional diversity of IDPs (surfaces of BCG-type asteroids = pyroxene-rich IDPs; surface of P- and D-type asteroids = mixture of pyroxene-rich and olivine-rich IDPs; interior of large BCG-, P- and D-types = hydrous IDPs?—see conclusion) whereas most comets (>65%; see A.6, A.7, Figures 6–8) are compatible with olivine-rich IDPs only. MBAs consequently appear as the most likely parent bodies of a large fraction of IDPs (and thus of most sporadic meteors). In addition, the relatively low maximum temperatures experienced during atmospheric entry by pyroxene-rich and hydrous IDPs, which implies that they originally had relatively circular, prograde orbits (asteroid-like), support an asteroidal origin for these particles (Sandford & Bradley 1989). However, it should be noted that comets still are likely to be a major source of olivine-rich IDPs.

Overall, our results imply that IDPs sample a larger fraction of MBAs (in terms of mass) than meteorites (Figure 3). In particular, whereas meteorites have been known for a long time to be representative of the surfaces of metallic, rocky and hydrated asteroids (∼34% of the mass of the belt), it now appears that only IDPs are representative of the surfaces of icy asteroids (∼66% of the mass of the belt). Note that a compositional overlap (and thus an overlap in the asteroid sampling) likely exists between the two classes of extraterrestrial material (e.g., between CI/CM meteorites and hydrous IDPs).

Zoom In Zoom Out Reset image size

Finally, combined with existing density measurements (Carry 2012; Marchis et al. 2012), our results provide additional evidence to the "comet-like" nature of BCG-type asteroids. In particular, they support the possibility that not only P- and D-type asteroids (Levison et al. 2009) but also a large fraction of the C-complex asteroids (namely BCG-type asteroids) are implanted bodies from the outer solar system, as proposed in the Grand Tack model (Walsh et al. 2011). A formation location closer to the Sun for BCG-type asteroids (pyroxene-rich) with respect to D-type asteroids (olivine-rich) as proposed in the Grand Tack model would suggest, following our results, that the abundance of olivine relative to pyroxene was higher in the outer part of the disk than in the inner part, in perfect agreement with the trend observed in young disks around other stars (Bouwman et al. 2008; Sargent et al. 2009). On the basis of both our current understanding of the early dynamical evolution of the solar system (Grand Tack model; Walsh et al. 2011) as well as previously established asteroid–meteorite links (Burbine 2014, pp. 365–415), our results suggest that meteorites likely sample bodies formed in the inner region of the solar system (0.5–4 AU) whereas IDPs sample bodies (both asteroids and comets) formed in the outer region (>5 AU). Our work thus opens the possibility that our IDP collections sample a large diversity of small bodies that formed in the outer solar system, thus ultimately enabling a better comprehension of the various processes that occurred in this region of the solar system during the earliest epochs.

Long considered to reflect their compositional diversity, meteorites have been used to decipher the origin, formation, and evolution of MBAs. However, finding a plausible meteoritic analog for low-density icy asteroids (i.e., for ∼2/3 of the mass of the asteroid belt) has been an unsuccessful quest for decades. It has thus been gradually accepted in the community that these objects, although they comprise most of the mass of the asteroid belt, are unsampled in our meteorite collections. Strangely, the interesting findings by Bradley et al. (1996) that opened the possibility that BCG-, P- and D-type asteroids could be linked to IDPs have never really found an echo in the community of asteroid spectroscopists. The same can be said regarding predictions by Dermott et al. (2002) based on observations of solar system dust bands that were advocating an asteroidal origin for most IDPs.

Here we show that BCG-, P- and D-type asteroids are in fact well sampled in our collections, namely as IDPs, thus confirming earlier suggestions/predictions (e.g., Bradley et al. 1996; Dermott et al. 2002 and references therein). To reach this conclusion, we rely on spectral data covering a large wavelength range (0.4–13 μm) thus significantly extending the coverage by Bradley et al. (1996). Our extended wavelength coverage further allows us to pinpoint the parent bodies of two important families of IDPs: BCG-type asteroids with pyroxene-rich IDPs, and P- and D-type asteroids with a mixture of pyroxene-rich and olivine-rich IDPs. Comets, on the other hand, appear mostly compatible with olivine-rich IDPs and to a lesser extent with a mixture of pyroxene-rich and olivine-rich IDPs. Overall, our results imply that IDPs sample a larger fraction of MBAs than meteorites. Reciprocally, they suggest that asteroids are the likely parent bodies of a large fraction of IDPs (as suggested earlier by Dermott et al. 2002), hence of a large fraction of sporadic meteors. Considering that current models of the early dynamical evolution of the solar system (Walsh et al. 2011) predict that icy asteroids are implanted bodies from the outer solar system, our findings thus imply that IDPs may also sample the building blocks of icy bodies in the outer solar system including giant planet satellites and TNOs and not only those of comets as previously suggested. Our work thus opens the possibility that our IDP collections sample a large diversity of bodies that formed in the outer solar system, ultimately enabling a better comprehension of the various processes that occurred over an extensive range of radial distances in the outer solar system (>5 AU) during the earliest epochs, hence nicely complementing the information we have for the inner solar system via the study of meteorites.

There are, however, several issues (listed below as questions) remaining that require critical new data and/or new modeling efforts.

• 1.  
  Which objects are the parent bodies of hydrous IDPs? Could slightly hydrated IDPs be associated with hydrated P-and D-type asteroids (whose spectra exhibit a 3 μm absorption band; Takir & Emery 2012)? or is the Tagish Lake meteorite a better analog for these objects? Also, since the mineralogy of hydrous IDPs is roughly similar but not identical to that of CI/CM chondrites (Bradley 1999 and references therein), it is a priori excluded that hydrous IDPs originate from Ch and Cgh-types. Instead, do hydrous IDPs originate from the interior of BCG-, P- and D-type asteroids where aqueous alteration has certainly occurred?
• 2.  
  Asteroid families likely are a main source of meteorites (Vernazza et al. 2008, 2014). Is this also the case for IDPs? That is, which objects contribute the most to the IDP flux and why?

Finally, the acquisition of near-infrared spectra (0.8–4 μm) for all classes of IDPs will help refine the associations proposed in this work.

We thank the referee for his pertinent and constructive remarks. This work has been partly carried out thanks to the support of the A*MIDEX project (n° ANR-11-IDEX-0001-02) funded by the "Investissements d'Avenir" French Government program, managed by the French National Research Agency (ANR).

A.1. Metamorphosed CMs as Improbable Analogues of BCG-type Surfaces

Metamorphosed CM chondrites, which represent ∼0.2% of the falls, have been proposed in the past as analogues of BCG-type (B-, C-, Cb- and Cg-type) surfaces (Hiroi et al. 1993) but such a possibility is presently untenable for three reasons.

• 1.  
  First, metamorphosed CM chondrites only represent ∼10% of all CM chondrites whereas BCG-types are twice as abundant as CM-like bodies (Ch and Cgh-types) (Rivkin 2012), implying that BCG-types would be ∼20 times less abundant than Ch and Cgh-types among meteorite falls. Whereas it is well known that atmospheric entry biases our meteorite collections by preferentially selecting the denser and more compact objects (which explains, for example, why ordinary chondrites are largely over-represented in our collections with respect to carbonaceous chondrites, such as CI and CM chondrites) it has been shown recently by Vernazza et al. (2014) that for materials having similar densities and tensile strength meteorite falls are representative of the compositional diversity of the asteroid belt within a factor of ∼4. This implies that metamorphosed CM chondrites should be at least as abundant as typical CMs in our collections; but they are not.
• 2.  
  Second, the high-precision density measurements collected so far for a few BCG-types indicate that these are significantly lower than those of Ch and Cgh-types (BCG-types (45) Eugenia, (90) Antiope, (253) Mathilde and (762) Pulcova have densities in the 0.8–1.5 g cm⁻³ range whereas Ch- and Cgh-types (41) Daphne, (121) Hermione and (130) Elektra have densities in the 1.9–2.4 g cm⁻³ range; Carry 2012; Marchis et al. 2012). Yet, the opposite trend would be expected for bodies that were heated throughout so that their near surface reached temperatures of the order of ∼500 °C. Instead, it appears that BCG-types have densities in the same range as P- and D-type asteroids (low densities which suggest that these bodies contain a high fraction of ice). What's more, water ice has been observed at the surface of some BCG-type asteroids (Campins et al. 2010; Rivkin & Emery 2010; Takir & Emery 2012), which is incompatible with the presence of heated CM material at the surface.
• 3.  
  Third, the mid-infrared spectral properties of heated CM chondrites (Figure 5) are at odds with those of BCG-type asteroids (which also applies to CI and CM chondrites).

In summary, there are no meteorites in our collections whose spectra match well those of BCG-type asteroids over both the 0.4–2.5 and 8–13 μm ranges.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

A.2. P- and D-type Surface Material Absent from Our Meteorite Collections

A.3. The "Comet-like" Nature of Low-density Asteroids

A.4. Observation and Data Reduction

A.5. Representativeness of Our Dataset

A.6. Pyroxene-rich and Phyllosilicate-rich IDPs as Improbable Analogues of Cometary Dust

In Figures 6 and 7, we compare the spectra of the different classes of IDPs with those of comets (including both long-period and short-period comets). It appears that more than 50% of the comets have spectral properties compatible with olivine-rich IDPs mixed with a minor fraction of pyroxene-rich IDPs (Figure 6). The remaining comets but one (Austin) are best fitted if a pyroxene-rich IDP at the ∼30%–50% level is included in the mixture. This is for example the case for comet Linear. Overall, the composition of cometary dust overlaps well with the one inferred for P- and D-type asteroids, although some comets (∼1/3) appear to be more olivine-rich than D-types (e.g., Hale–Bopp, Halley). In our sample, there is one clear exception (in terms of composition), namely comet Austin for which a pyroxene-rich IDP provides the best fit. However, one needs to remain cautious about this comet' spectrum because of its very low signal-to-noise ratio. There are a few other comet spectra in the literature that are not displayed here. In all cases but one (comet Wilson, Hanner et al. 1994), a clear olivine feature at 11.3 μm is observed preferentially linking these comets to IDPs enriched in olivine.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Finally, we use Figure 7 to show how poorly both pyroxene-rich and phyllosilicate-rich IDPs fit cometary spectra, implying that these IDPs cannot originate from comets.

A.7. Modeling of the Spectra with a Radiative Transfer Code

A.7.1. The Model

We applied a radiative transfer model in both the 0.4–2.5 and 8–13 μm ranges to analyze the mineralogies of BCG-type asteroids, as well as the composition of P- and D-types and comets (for comets we used the 8–13 μm range only). In the visible and near-infrared spectral range, the single-scattering albedo was calculated using the Mie theory (Bohren & Huffman 1983) for grains with sizes small or comparable to the wavelength, and from optic-geometric equations for larger grains. The single-scattering albedo was then injected in Hapke's equation of the bidirectional reflectance (Hapke 1993). In the mid-infrared range, the emissivity was approximated by the absorption efficiency factor (${Q}_{\mathrm{abs}})$ calculated for a distribution of hollow spheres with small sizes (<1 μm) compared to the wavelength (Min et al. 2003).

We used the optical constants (n and k) of common end-member minerals retrieved from the Jena database (http://www.astro.uni-jena.de/Laboratory/Database/databases.html) found in anhydrous chondritic porous IDPs (Bradley 2005). Chondritic porous IDPs are heterogeneous aggregates of predominantly submicrometer-sized crystalline mineral grains (olivine, pyroxenes, iron-rich sulfides, and FeNi metal), polycrystalline aggregates (olivine-like and enstatite-like), silicate glasses (GEMS), and carbonaceous material (Bradley 2005). Enstatite and forsterite with <5% mol. Fe are the most common crystalline silicates, although crystals with up to 30% mol. Fe are also observed. Sulfides are the second most abundant class of minerals in chondritic porous IDPs, the most common sulfide being pyrrhotite with grain sizes in the range ∼10 nm–5 μm.

The free parameters of our model are the relative abundance of the components (whose sum must be equal to 1), the mineral chemistries of the various components (variation of the Mg-number) and the grain size distribution (between 0.2 and 1 μm). Finally, an IDL routine was used to find the minimum rms residual between the measured spectrum and the computed spectrum.

A.7.2. Results

A.7.2.1. 0.4–2.5 μm Range

In the 0.4–2.5 μm range, silicate glasses and sulfides were necessary end-members for reproducing the spectral properties of BCG-type asteroids (in particular for reproducing the broad band whose center varies from ∼1 to ∼1.3 μm). Since both enstatite and forsterite possess a neutral, flat spectrum in this wavelength range (due to the small grain size), they were not necessary components to achieve good fits of BCG-types. Amorphous carbon helped improving the fits of BCG-type spectra in the visible part of the spectrum. For P- and D-types, the use of a space-weathering model (Brunetto et al. 2006) was necessary for reproducing their high spectral slopes.

Overall, it appears that the spectra of BCG-, P- and D-type asteroids are well reproduced by our model if we assume end-members and grain sizes similar to those observed in chondritic porous IDPs. The main issue in this wavelength range is the lack of absorption feature in both enstatite and forsterite spectra which prevents us from a detailed compositional analysis. In the case of surfaces covered by small grains (smaller than 1 μm), the mid-infrared spectral range is known to be significantly more powerful for the determination of the composition of silicate assemblages than the 0.4–2.5 μm range alone. We therefore used, as a next step, available spectra of BCG-type, P- and D-type asteroids and comets in the 8–13 μm range to constrain their surface composition in more detail.

A.7.2.2. 8–13 μm Range

In the mid-infrared, crystalline silicates (both olivine and pyroxene) were essential ingredients for reproducing the spectral properties—in particular the peaks—of BCG-type and P/D-type asteroids (as well as those of comets). Our spectral analysis thus implies that both olivine and pyroxene (both in amorphous and crystalline phase) are present at the surfaces of BCG-type asteroids as well as those of P- and D-types and comets (Figure 8). In terms of relative abundance, we found that BCG-type asteroids are more pyroxene-rich than P-types, which are themselves more pyroxene-rich than D-types and comets (the latter being conversely more olivine-rich). This finding is supported by a direct comparison of IDP spectra with both asteroid and comet spectra (Figures 1, 8).

Zoom In Zoom Out Reset image size