A NanoSIMS and Raman spectroscopic comparison of interplanetary dust particles from comet Grigg-Skjellerup and non-Grigg Skjellerup collections


  • Jemma DAVIDSON,

    1. Planetary and Space Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, UK
    2. Present address: Hawai’i Institute of Geophysics and Planetology, University of Hawai’i at Manoa, Honolulu, Hawai’i 96822, USA
    Search for more papers by this author
  • Henner BUSEMANN,

    1. Planetary and Space Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, UK
    2. School of Earth, Atmospheric and Environmental Sciences, The University of Manchester, Oxford Road, Manchester M13 9PL, UK
    Search for more papers by this author
  • Ian A. FRANCHI

    1. Planetary and Space Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, UK
    Search for more papers by this author

Corresponding author. E-mail: davidson@higp.hawaii.edu


Abstract– Interplanetary dust particles (IDPs) are the most primitive extraterrestrial material available for laboratory studies and may, being likely of cometary origin, sample or represent the unaltered starting material of the solar system. Here we compare IDPs from a “targeted” collection, acquired when the Earth passed through the dust stream of comet 26P/Grigg-Skjellerup (GSC), with IDPs from nontargeted collections (i.e., of nonspecific origin). We examine both sets to further our understanding of abundances and character of their isotopically anomalous phases to constrain the nature of their parent bodies. We identified ten presolar silicates, two oxides, one SiC, and three isotopically anomalous C-rich grains. One of seven non-GSC IDPs contains a wealth of unaltered nebula material, including two presolar silicates, one oxide, and one SiC, as well as numerous δD and δ15N hotspots, demonstrating its very pristine character and suggesting a cometary origin. One of these presolar silicates is the most 17O-rich discovered in an IDP and has been identified as a possible GEMS (glass with embedded metal and sulfides). Organic matter in an anhydrous GSC IDP is extremely disordered and, based on Raman spectral analyses, appears to be the most primitive IDP analyzed in this study, albeit only one presolar silicate was identified. No defining difference was seen between the GSC and non-GSC IDPs studied here. However, the GSC collectors are expected to contain IDPs of nonspecific origin. One measure alone, such as presolar grain abundances, isotopic anomalies, or Raman spectroscopy cannot distinguish targeted cometary from unspecified IDPs, and therefore combined studies are required. Whilst targeted IDP populations as a whole may not show distinguishable parameters from unspecified populations (due to statistics, heterogeneity, sampling bias, mixing from other cometary sources), particular IDPs in a targeted collection may well indicate special properties and a fresh origin from a known source.


Interplanetary dust particles (IDPs) are complex, heterogeneous aggregates of primitive solar system materials, typically 5–25 μm in diameter, although cluster IDPs (large IDPs that fragment on impact with the collector flag creating many smaller particles) can be >100 μm (Bradley 2003). IDPs were routinely collected in the stratosphere by NASA high-altitude research aircraft (Brownlee 1985; Sandford 1987). Most IDPs can be broadly grouped into two morphological types: chondritic porous (CP), consisting primarily of anhydrous minerals (e.g., forsterite, enstatite, glass with embedded metal and sulfides [GEMS], Fe-sulfides, and carbonaceous material), and chondritic smooth (CS), consisting primarily of hydrated layer lattice silicates, with minor components of carbonates, sulfides, carbonaceous material, and some anhydrous crystalline silicates (Bradley 2003; Floss et al. 2006). The latter, also known as hydrated IDPs, have been speculated to be of asteroidal origin (Bradley and Brownlee 1986; Keller et al. 1992; Rietmeijer 1992), while the anhydrous, or CP IDPs, are thought to be of cometary origin (e.g., Brownlee et al. 1995; Bradley 2003). Throughout this paper the term “primitive” refers to pristine samples that have experienced minimal secondary processing (they may be aqueously but not thermally altered), have disordered C, and contain isotope anomalies. As shown with the organic matter in meteorites, aqueous alteration does not alter the degree of order of C (e.g., Busemann et al. 2007) and does not readily destroy isotope anomalies in H or N, carried by organic matter (e.g., Alexander et al. 2007). The primitive nature of anhydrous CP IDPs is reflected by their high presolar grain abundances (e.g., Messenger et al. 2005; Floss et al. 2006, 2010; Busemann et al. 2009a) and profuse isotopic anomalies; IDPs frequently exhibit enrichments in deuterium (e.g., Zinner et al. 1983; McKeegan et al. 1987; Messenger et al. 1995) and 15N (e.g., Stadermann et al. 1989; Stadermann 1991; Messenger et al. 2003a) that have been interpreted as the result of either processing in the interstellar medium predating the formation of the proto-planetary disk (e.g., Messenger et al. 2003a), or ion-molecule chemistry in the coldest regions of the protoplanetary disk (e.g., Aléon 2010, and references therein).

Generally, the origins of IDPs remain unknown although some have been tentatively linked to specific parent bodies (e.g., Busemann et al. 2009a; Floss et al. 2010). Since comets, which spend most of their lifetimes in the cold outer regions of the solar system, contribute significantly to the interplanetary dust population crossing the Earth at 1 AU (Messenger 2002), many IDPs must have cometary origins (e.g., Nesvorný et al. 2010). However, it is difficult to unambiguously establish cometary origins for specific IDPs, particularly since dust returned by the Stardust mission, from the Jupiter-family comet 81P/Wild 2, was unexpectedly found to contain material processed in the inner solar system (Brownlee et al. 2006, 2012; Ishii et al. 2008). This may indicate that the original idea of a dichotomy between outer solar system comets and inner solar system asteroids should more realistically be replaced by a continuum of planetary bodies with carbonaceous chondrites being the intermediate link (e.g., Gounelle et al. 2006; Aléon et al. 2009).

Recently, IDP collections have been targeted towards those that coincide with the passage of Earth through the dust trails of specific comets, such as comet 26P/Grigg-Skjellerup (the corresponding IDPs were examined, e.g., in Busemann et al. 2009a) and comet 55P/Tempel-Tuttle (see, e.g., Floss et al. 2010). Messenger (2002) identified comet 26P/Grigg-Skjellerup as one of four short-period (Kuiper belt) comets that had, in the past century, Earth-crossing dust streams and low Earth-encounter velocities. Messenger (2002) predicted that when comet Grigg-Skjellerup’s dust stream passed through Earth’s atmosphere in April 2003 between 1 and 50% of the total IDP flux would originate from the comet. As a result, NASA flew two dedicated large area stratospheric collectors (L2054 and L2055), and two small collectors (U2120 and U2121) on a NASA ER-2 aircraft for a total of 7.9 h on April 30, 2003, over the Southwestern USA (see http://www-curator.jsc.nasa.gov/dust/Leonid_shwr_Grigg_Skjellerup.cfm). Two additional large area collectors (L2052 and L2053) were also flown despite having already accumulated IDPs of nonspecific origins during previous flights.

Since collection, a small number of IDPs from the Grigg-Skjellerup collection (GSC) have been shown to be extremely primitive in nature (Nittler et al. 2006; Nguyen et al. 2007a; Nakamura-Messenger et al. 2008, 2010; Busemann et al. 2009a). Busemann et al. (2009a) speculated that some GS collection IDPs they analyzed may be associated with comet Grigg-Skjellerup because of their particularly pristine character. Floss et al. (2010) also reported the presence of abundant presolar and other isotopic anomalies in an IDP from the 55P/Tempel-Tuttle targeted stratospheric dust collector suggesting a cometary origin. To test the hypothesis that the targeted cometary collections contain more abundant pristine material than nontargeted collections we compared IDPs from the comet Grigg-Skjellerup collection and untargeted collections. This study aims to further our understanding of the abundances, characteristics, and distribution of their isotopically anomalous phases and presolar grains to constrain their nature and that of their parent bodies.

Analytical Procedure

Sample Preparation and Characterization

IDPs of anhydrous fine-grained appearance (as requested) were picked from collector flags (Table 1) at the Johnson Space Center’s Cosmic Dust Laboratory and placed on glass dimple slides upon which microRaman spectroscopic analyses (see below) were undertaken. Three designated IDPs (L2006 U8 cluster #3, W7029 N1 cluster #2, W7029 N2 cluster #2) could not be relocated on their glass slides. The IDPs were transferred to high-purity gold foils (mounted on aluminum stubs) using an optical microscope and a tungsten needle attached to a Narishige micromanipulator and then pressed using spectroscopic grade sapphire windows and a custom sample press.

Table 1. Names and characteristics of the interplanetary dust particles studied here.
NameaCollector/particle no.ClusterSize (μm) BPbSize (μm) APcArea (μm2)cComp.d
  1. aIDPs were assigned alternative names after explorers from the 1400s to 1600s.

  2. bOriginal size before pressing (BP) into Au foil (i.e., size at time of microRaman analysis).

  3. cSize after pressing (AP) into Au foil (i.e., size at time of NanoSIMS analysis). *No size is given for Pizarro as it fragmented into 4 pieces.

  4. dCP = chondritic porous, CS = chondritic smooth (Bradley 2003).

  5. eIndicates that Polo appears to be a CS fragment of CP cluster particle L2005 12.

Non-GSC cluster IDPs 392
PizarroL2005 AR83111 × 11*64CP
CortesL2005 AR931 7 × 813 × 1196CP
MagellanL2005 AR1031 7 × 912 × 13109CP
ColumbusL2005 AR1112 4 × 4 6 × 618CP
PoloL2005 AR1212 6 × 5 5 × 514CP/CSe
HudsonL2006 U93 7 × 8 9 × 1067CP
RaleighL2006 U103 8 × 5 8 × 524CP
GSC cluster IDPs 259
HawkinsL2054 P44 8 × 8 9 × 912CS
DrakeL2055 M3510 × 1216 × 17173CP
FrobisherL2055 M411 8 × 912 × 1074CS

The IDPs were characterized using a Zeiss Supra 55VP field emission scanning electron microscope (FE-SEM) at the Open University (OU); SEM-based energy dispersive X-ray (EDX) analysis was used to determine whether the IDPs had chondritic abundances of major rock-forming elements such as Si, Mg, Fe, S, Ca, and Al. EDX spectra were taken at a central point on each IDP (typical spot size approximately 2 μm) using an accelerating voltage of approximately 15 kV (working distance of 10 mm). All non-GSC IDPs and GSC IDP Drake had chondritic spectra (Davidson 2009), GSC IDP Hawkins had a large Al-peak resulting from the Al2O3 grain attached to the smaller chondritic particles (Fig. 1h), whilst GSC IDP Frobisher had a large Al-peak apparently inherent to the sample. A further two particles from GSC collector U2121 (H4 and H5 from clusters #4 and #5, respectively) were found to be most likely terrestrial in nature (high Cd counts suggest that they are paint) and are not considered further. Secondary electron (SE; Fig. 1) images were acquired (4 kV at a working distance of 5 mm) both prior to and after NanoSIMS ion microprobe analysis.

Figure 1.

 Secondary electron (SE) images of (a–g) non-GSC (Grigg-Skjellerup collection) and (h–j) GSC IDPs studied (Davidson 2009). (a) Pizarro fragmented into four pieces (“a, b, c, d”) during picking. (a-c) are from the same cluster, as are (d, e), and (f, g); (h, i, j) are all from separate cluster particles. (h) Hawkins contains a large aluminum oxide (Al2O3; outlined) which is attached to several small chondritic (“C”) fragments (also outlined). All scale bars are 2 μm.

The three extraterrestrial particles obtained from the Grigg-Skjellerup Collection (GSC) represent three separate cluster particles (Table 1), providing a significant addition to seven cluster particles and three individual particles analyzed to date (Busemann et al. 2009a; Floss et al. 2010; and references therein).

MicroRaman Spectroscopy

MicroRaman spectroscopy is a relatively nondestructive technique that requires minimal sample preparation. Laser Raman analyses were conducted with a Horiba Jobin Yvon Labram HR Raman system at the OU (see, e.g., Rotundi et al. [2008] for further details). Excitation was delivered by an argon ion laser (514.5 nm) and the spectral resolution is 3 Δcm−1 using a 600 gr/ mm−1 grating. To prevent thermal modification of the IDPs, the laser delivered a power of approximately 0.06 mW at the sample surface. The beam was focused with a ×50 long working distance objective giving a spatial resolution of approximately 2 μm. All microRaman analyses were undertaken prior to NanoSIMS analysis known to induce amorphization of the top layers of an analyzed sample due to high-energy Cs+ ion implantation (e.g., Busemann et al. 2007, 2009a). Spectra were accumulated across each particle with 2 μm steps in both x and y, as 5 sets of 30-s integrations. The carbonaceous D (breathing mode only active in “disordered” carbon at approximately 1355 cm−1, e.g., Ferrari and Robertson 2000) and G (“graphite” band at approximately 1581 cm−1 due to the stretching of neighboring sp2 atoms) band features (band intensity I, position ω, and full width at half maximum Г) were fitted in the range between approximately 850 and 2100 cm−1 to Lorentzian profiles for the Raman bands with a free-floating linear background to accommodate the steep sloping fluorescence baseline (see Busemann et al. [2007] for selection criteria and data reduction procedures). No spectroscopic features other than the D and G bands were detected due to the large Raman response of carbon.

NanoSIMS Ion Imaging

Three sets of isotopic measurements were undertaken using the NanoSIMS 50L ion microprobe at the OU in order to determine the spatial distributions of the H, C, N, O, and Si isotopes in all IDPs. Additionally 24Mg16O (mass-40) was mapped to distinguish SiC grains from other isotopically anomalous phases. IDPs were presputtered with a primary beam of 16 keV Cs+ ions and probe currents of typically approximately 100 pA for up to ten minutes (depending on fragment size) until sputter equilibrium was reached. All measurements were made in scanning imaging mode, where the primary Cs+ beam is rastered over the sample and secondary ions are simultaneously collected in up to seven electron multiplier detectors. Charge compensation was not necessary. Frame sizes of 256 × 256 or 512 × 512 pixels were used for all images depending on size of the IDP being analyzed (areas ranged from 7 × 7 to 17 × 17 μm2). The primary beam of 16 keV Cs+ ions (and hence approximate spatial resolution) was typically approximately 100 nm for C, N, O, and Si, and approximately 200 nm for H isotope measurements. Data were reduced using the L‘IMAGE software package (L. R. Nittler, Carnegie Institution of Washington), they were not corrected for QSA (e.g., Slodzian et al. 2004) as this should not have a significant effect on the ratios determined here with low primary ion currents. Variable numbers of image planes were collected for each different set of measurements (see below). Stage shift of typically one to three pixels (approximately 50–150 nm) during analysis was corrected during data reduction. Grains and OM were considered presolar or anomalous if their relevant isotopic compositions (e.g., either 16O/17O or 16O/18O for presolar silicates and oxides, 12C/13C for presolar SiC and other C-anomalous grains, and D/H for OM) were >3σ from the average isotopic composition of each IDP.

1. In order to detect presolar SiC, graphite, anomalous-C grains, and other organic matter (OM; Table 2), C and N isotopes were measured as negative ions (12C, 13C, 12C14N, 12C15N, 16O, 28Si, and 28Si12C; Table 3) at a mass resolution of m/Δm = approximately 9000 (all values given in this work according to CAMECA definition), sufficient to resolve all interferences from neighboring peaks (e.g., 10B16O from 12C14N and 11B16O from 12C15N; see Zinner et al. 1989; Hoppe et al. 1995). A 16 keV Cs+ primary ion beam and current of 1 pA was used for the measurements. Each image consists of six planes (total integration time of approximately 70 min). Orgueil IOM (characterized by Alexander et al. 2007) was measured before and after analysis of IDPs to act as an isotopic standard (for C and N) to correct for instrumental mass fractionation.

Table 2. Numbers and abundances of presolar grains and other isotopically anomalous regions in IDPs.
NameNumber of presolar grainsOther isotopic anomalies
(#, Area%)(#, Area%)
  1. aAbundance errors are 1σ after Gehrels (1986).

Non-GSC cluster IDPs
a1  1  13.9
b  1     
c      22.2
d  1 16.631.3
Pizarro (all)1 2113.261.6
Cortes   1  80.9
Magellan  1 11.421
Columbus     111.416.7
Polo   1 112.411.4
Hudson 14   71.5
Raleigh  1 33031.7
Total (all non-GSC)119272.5281.2
Abundancesa70 ± 60100 ± 802800 ± 900800 ± 500    
GSC cluster IDPs
Drake  1   32.7
Total (Drake)  1  3
Abundancesa  920 ± 700  2.7
Table 3. Carbon and nitrogen isotopic compositions of the IDPs studied here, with isotopically anomalous C grains (SiC and other) and anomalous nitrogen regions.
NameArea 12C/13Caδ13C (‰)a 14N/15Naδ15N (‰)aArea (μm2)Diameter (nm)
  1. aErrors are 1σ.

  2. bBulk composition is based on analysis of Pizarro fragment “a” (Fig. 1a), which accounts for approximately 20% of the total area of Pizarro.

Non-GSC cluster IDPs
aBulk88.4 ± 0.67 ± 7.4219.2 ± 1.7243 ± 9.5154430
aSiC9236 ± 62−623 ± 99378.7 ± 122−281 ± 2310.03200
aN285.5 ± 1.441 ± 17146.8 ± 1.2856 ± 160.6870
cN297.1 ± 1.6−83 ± 15176.3 ± 4.4546 ± 390.22530
cN389.4 ± 4.4−5 ± 49157.7 ± 5.9728 ± 650.16430
dN297.0 ± 2.2−83 ± 21156.8 ± 4.8737 ± 530.13400
dN378.2 ± 3.0137 ± 43184.8 ± 10.2475 ± 820.07300
dN478.2 ± 2.0138 ± 28185.3 ± 7.6471 ± 600.13400
CortesBulk87.7 ± 0.415 ± 4.5276.4 ± 0.9−14 ± 310111350
 N295.5 ± 4.3−68 ± 42134.0 ± 5.81034 ± 890.08310
 N387.3 ± 2.219 ± 26170.4 ± 8.1599 ± 760.18480
 N485.8 ± 3.238 ± 39168.6 ± 8.1617 ± 770.12390
 N584.1 ± 4.358 ± 54156.4 ± 9.9742 ± 1100.08320
 N690.7 ± 8.4−19 ± 91131.1 ± 13.41078 ± 2130.06270
 N788.9 ± 3.51 ± 40189.0 ± 9.6442 ± 740.15430
 N987.4 ± 3.718 ± 43178.2 ± 9.9529 ± 850.11400
 N1088.0 ± 2.511 ± 29182.1 ± 11.1497 ± 910.09300
MagellanBulk87.8 ± 0.214 ± 2.1268.4 ± 0.714.2 ± 311312000
 N290.5 ± 1.0−17 ± 11134.9 ± 3.91020 ± 590.7900
 N388.3 ± 1.48 ± 16145.2 ± 3.5876 ± 450.47800
ColumbusBulk87.0 ± 0.423 ± 4.9272.5 ± 2.1−1 ± 8205000
 N280.8 ± 1.3101 ± 18136.5 ± 3.6996 ± 520.13400
PoloBulk88.0 ± 0.712 ± 8.3327.1 ± 3.0−167 ± 8184800
 N290.4 ± 2.8−16 ± 31242.9 ± 9.0122 ± 420.25600
HudsonBulk87.1 ± 0.222 ± 2.7266.5 ± 0.622 ± 3739700
 C3112.2 ± 6−207 ± 41290.0 ± 25−60 ± 800.04200
 N299.9 ± 4.2−109 ± 37140.7 ± 8.9936 ± 1220.08300
 N385.9 ± 1.936 ± 23184.5 ± 4.9477 ± 400.09300
 N484.6 ± 1.851 ± 22189.9 ± 5.0435 ± 380.12400
 N589.7 ± 2.0−8 ± 22195.0 ± 8.5397 ± 610.1400
 N680.4 ± 3.6107 ± 49150.3 ± 7.8812 ± 940.08300
 N784.1 ± 3.459 ± 42162.3 ± 9.6679 ± 990.08300
 N989.7 ± 1.3−7 ± 14189.6 ± 3.7437 ± 280.56800
RaleighBulk87.6 ± 0.416 ± 4.5278.7 ± 3.0−22 ± 11296100
 N288.2 ± 2.99 ± 33192.0 ± 9.0419 ± 660.15400
 N585.1 ± 2.346 ± 28184.1 ± 7.0480 ± 560.23500
 N682.2 ± 3.183 ± 41184.0 ± 11.2481 ± 900.11400
GSC cluster IDPs
HawkinsBulk87.2 ± 1.221 ± 13.5336.5 ± 11−190 ± 27437400
DrakeBulk89.1 ± 0.1−1 ± 1.5219.9 ± 0.4239 ± 217715000
 N387.2 ± 3.321 ± 39148.5 ± 5.7835 ± 700.17500
 N796.4 ± 5.0−77 ± 48166.8 ± 6.0633 ± 580.21500
 N987.4 ± 4.418 ± 51165.2 ± 7.9649 ± 790.09300
FrobisherBulk87.0 ± 1.123 ± 13.3249 ± 7.394 ± 32769800

2. In order to detect presolar silicates and oxides, O and Si isotopes were measured as negative ions (16O, 17O, 18O, 28Si, 29Si, 30Si, with 28Si12C to distinguish silicate from nonsilicate grains; Table 4) at a mass resolution of m/Δm = 9000 (CAMECA definition), required to resolve the 17O peak from the 16OH peak. Although it was not possible to resolve the 28Si12C peak from 24Mg16O, collection of mass-40 data permits the tentative identification of presolar SiC grains and correlation with those positively identified during C and N measurements. Currents of approximately 1 pA were used. Images consisted of typically 20 frames with an integration time of 4.5 h. Oxygen isotope standards were not measured because the bulk O isotopic compositions of IDPs are solar to within 2% (Aléon et al. 2009; Nakashima et al. 2012) and this study aimed to locate presolar grains only. The average O isotopic compositions of each IDP were taken as solar (SMOW) values (16O/17O = 2625, 16O/18O = 499). Silicon isotopes were also internally normalized assuming solar compositions.

Table 4. Oxygen and silicon isotopic compositions of O-rich presolar grains in the IDPs studied here.
Name Grainaδ17O (‰)b 16O/17Obδ18O (‰)b 16O/18Obδ29Si (‰)b,cδ30Si (‰)b,c 28Si/16OArea (μm2)dDiameter (nm)d
  1. aMineralogy listed here is based on the 28Si/16O and 24Mg16O/16O ratios from the NanoSIMS data (“Sil” = silicate, “Ox” = oxide).

  2. bErrors are 1σ.

  3. cSilicon isotopic compositions are only reported for presolar silicate grains

  4. dDetermined from NanoSIMS O isotope maps.

Non-GSC cluster IDPs
aOx5106 ± 392362 ± 85−5 ± 14501 ± 7  0.010.103360
bSil5015715 ± 435156 ± 4−184 ± 11611 ± 9−35 ± 17−28 ± 230.020.155440
dSil379 ± 172420 ± 3850 ± 8475 ± 4−7 ± 9−10 ± 110.020.253570
CortesOx19372 ± 172436 ± 38107 ± 8451 ± 3  0.010.209520
Magellan Sil2420 ± 331839 ± 42−13 ± 13505 ± 7−14 ± 126 ± 160.040.096350
PoloSil2411 ± 301851 ± 399 ± 9494 ± 422 ± 1235 ± 150.020.09340
HudsonSil2318 ± 451982 ± 68−37 ± 10518 ± 58 ± 1324 ± 130.030.106370
 Sil3296 ± 412016 ± 64−39 ± 13519 ± 727 ± 1513 ± 180.030.068300
 Sil4258 ± 302075 ± 50−41 ± 12520 ± 723 ± 1842 ± 180.020.09340
 Sil5123 ± 292325 ± 591 ± 11498 ± 58 ± 115 ± 130.040.07300
RaleighSil2577 ± 241656 ± 25−19 ± 7508 ± 479 ± 1175 ± 120.020.175470
GSC cluster IDPs
DrakeSil2270 ± 492056 ± 79−47 ± 18523 ± 108 ± 1510 ± 180.040.159450

3. Hydrogen isotope measurements were the last to be made in all IDPs as their detection caused the most severe sample alteration/sputtering. Negative ions (1H, 2H, 13C, 18O) were measured at a mass resolution of m/Δm= 5000 using a primary ion current of 7–8 pA. Hydrogen and deuterium were measured to locate D/H anomalous regions, whilst C and O were analyzed to correlate with previous NanoSIMS images and characterize the carrier phases. Images consisted of five frames with a typical integration time of 15 min. Well characterized terrestrial (CHO; a lipid with the chemical formula C30H50O; e.g., Busemann et al. 2006a) and extraterrestrial samples (IOM; insoluble organic matter from Orgueil, GRO 95577, and EET 92042; Busemann et al. 2006a; Alexander et al. 2007) were analyzed as isotopic standards to correct for instrumental mass fractionation (Table 5). IOM samples were chosen as standards because material with large N- and D-enrichments were required to compare with the large potential enrichments expected in IDPs (i.e., nonterrestrial material)—to avoid the effects of hotspots averages of large areas of “bulk” material were taken. Standards were measured at regular intervals during measurements of the IDPs (approximately every three samples and at least twice a day). H isotopic compositions are reported relative to standard mean ocean water (SMOW; D/H = 0.00015576; Lodders and Fegley 1998). C/H ratios are reported here as ion ratios C/H and, hence, are rough estimates of the true C/H in bulk IDPs. IDPs are conglomerates of many different minerals at submicrometer scale, rendering calibration with standards difficult. While differing matrix chemical compositions and surface conditions create different ionization probabilities, we assume that on average the effects will be the same for all fine-grained IDPs in this study. The comparison of C/H in the IDPs in this study is possible as they were examined under the same experimental conditions.

Table 5. Bulk and hotspot/coldspot hydrogen isotopic compositions of the IDPs.
Name AreaD/HSMOWa,bδDSMOW (‰)b 12C/H− cSize (μm2)dDiam. (μm)d
  1. aD/H data are normalized to SMOW (standard mean ocean water; D/H = 0.00015576; Lodders and Fegley 1998).

  2. bErrors are 1σ.

  3. cErrors (1σ) are typically ±10%.

  4. dDetermined from NanoSIMS D/H isotope ratio maps.

  5. eCHO = terrestrial standard (lipid) with composition: C30H50O.

  6. fOrgueil IOM measured data are based on IMF corrections determined from terrestrial standard CHO and almost perfectly agree with the published value (Alexander et al. 2007) to within 2σ.

  7. gOrgueil IOM data from Alexander et al. (2007).

Non-GSC cluster IDPs
aBulk1.43 ± 0.06428 ± 62 93.3
dBulk1.30 ± 0.05297 ± 461.483.3
CortesBulk1.56 ± 0.04560 ± 411.08110.2
MagellanBulk1.59 ± 0.03593 ± 331.17810
 H24.53 ± 0.393528 ± 394 1.081.17
ColumbusBulk2.53 ± 0.081525 ± 821.173
 H24.66 ± 0.233655 ± 227 0.811.01
PoloBulk1.86 ± 0.08862 ± 820.652.4
 H23.59 ± 0.302590 ± 302 0.580.86
HudsonBulk1.68 ± 0.05678 ± 531.0679.3
RaleighBulk2.79 ± 0.061787 ± 620.9134
 H23.21 ± 0.302211 ± 298 0.530.82
 H34.72 ± 0.203718 ± 197 1.571.42
 H43.63 ± 0.172632 ± 169 1.621.43
GSC cluster IDPs
HawkinsBulk1.60 ± 0.09599 ± 850.3144.2
DrakeBulk2.14 ± 0.061141 ± 591.817114.8
FrobisherBulk1.65 ± 0.04651 ± 410.4709.5
“CHO”Bulk −152.8   
Orgueil IOM measuredfBulk1.95 ± 0.01948 ± 22   
Orgueil IOM publishedgBulk1.97972   

The potential impact of electron beam analysis on D/H compositions of samples has been noted by De Gregorio et al. (2010). To minimize impact on the samples analyzed here short exposure times with minimal current were used; the effects reported by De Gregorio et al. (2010) followed exposure to TEM conditions with an electron beam operating at 200 keV.


Sample Descriptions and General Observations

A total of ten IDPs were studied; three from the GSC plates L2054 and L2055 and seven not associated with any known extraterrestrial dust stream (“non-GSC” IDPs). All extraterrestrial fragments are from cluster IDPs and have been designated alternative names (see Table 1). Non-GSC and GSC IDPs represent three different clusters each. Cluster IDPs were requested as they are generally thought to be more primitive than individual IDPs based on the wider range of H and N anomalies seen in cluster IDPs (Messenger 2000; Messenger et al. 2003b), although this has since been disputed by Floss et al. (2006) and Busemann et al. (2009a).

Scanning electron microscopy (SEM), combined with C/H ratios determined by NanoSIMS analysis (Table 5), demonstrate that all non-GSC IDPs studied here (except for IDP Polo) are fine-grained, low-density, CP IDPs (Figs. 1a–d, 1f, and 1g). EDX analyses show that these IDPs are dominated by Mg-rich silicates. IDP Polo (Fig. 1e) appears to be very coarsely grained and not entirely CP-like. The C/H ratio of IDPs may act as an indicator of their nature; Aléon et al. (2001) identified different types of IOM in IDPs with C/H values ranging from 1.0 to 3.0, with lower C/H values associated with an additional source of C-free H considered to originate from phyllosilicate hydroxyl groups. Therefore, the C/H ratio of 0.6 determined here for Polo also indicates that this is a somewhat more hydrated grain, while the other CP grains had more typical CP-IDP C/H ratios of 0.9–1.8 (Table 5). Errors for C/H ratios are typically ±10% at 1σ. However, C/H ratios should be considered carefully as it is not possible to compare C/H ratios between different studies without calibration and metal-rich IDPs often show low C/H (e.g., Spring et al. 2010) without being classified as necessarily hydrous, and IDP Columbus (Fig. 1d), from the same cluster (L2005 12), is very finely grained, indicating that Polo represents a more coarsely grained, probably more hydrated, fragment of the original IDP. In the GS collection (Figs. 1h–j), Drake (Fig. 1i) appears to be a fine-grained, low-density, anhydrous (C/Hbulk∼1.8) IDP, whilst Hawkins (Fig. 1h) and Frobisher (Fig. 1j) are less porous, and have very low C/Hbulk ratios (0.3 and 0.4, respectively) indicating they have experienced more alteration and might be less primitive (either hydrous chondritic smooth, CS, IDPs; Bradley 2003; or “metal-rich”). Their bulk H-isotopic compositions (600 ± 90‰ and 65 ± 40‰, respectively, Table 5) confirm their extraterrestrial nature but also their lower degree of “primitiveness” compared to Drake (δD = 1140 ± 60‰). EDX analysis indicates that Drake is a Mg-rich silicate, whilst Hawkins and Frobisher are dominated by aluminum oxide, in line with their low C/H. IDP Hawkins (Fig. 1h) consists of two small chondritic particles attached to a large (8 by 4 μm), Al2O3 grain. Busemann et al. (2010) reported that another IDP (M2) from this cluster particle (L2054 cluster #4) was also composed of Al2O3, and lacked the spherical shape characteristic of rocket fuel additives commonly found in the stratosphere (Brownlee et al. 1976). A further two aluminum oxide IDPs (L2, I4) from the same cluster particle have been reported to have porosity similar to typical anhydrous IDPs (Nakamura-Messenger et al. 2008). However, the chondritic fragments attached to the Al2O3 grain studied here (Fig. 1h) appear to be more hydrous than those reported by Nakamura-Messenger et al. (2008); therefore, some areas of Hawkins may be hydrous fragments of an otherwise anhydrous parent cluster particle. The Al2O3 component of Hawkins has subsequently been analyzed for high-precision oxygen isotopes and appears to be a FUN-like inclusion similar to those seen in CAIs (Starkey et al. 2012). Further discussion of IDP Hawkins refers to the small, smooth chondritic particles attached to the Al2O3 grain only. Because Frobisher appears to be hydrated (it has a smooth appearance similar to other CS IDPs, is not metal-rich, and has a C/H ratio of 0.4), it is possible that this particle is of asteroidal origin (e.g., Bradley and Brownlee 1986; Keller et al. 1992; Rietmeijer 1992).

Presolar Grains

One way to assess the primitive nature of any extraterrestrial sample is to investigate its presolar grain components and their abundances. Abundances are calculated based on the area of presolar grains found compared to the area of analyzed sample, errors are based on counting statistics (after Gehrels 1986) as employed by other studies (e.g., Floss et al. 2006). Presolar silicates are the most abundant presolar grain type found in meteorites (e.g., Floss and Stadermann 2008, 2009a; Nguyen et al. 2010) with the exception of nanodiamonds, which may not all be of presolar origin; e.g., Dai et al. (2002), and in IDPs (e.g., Messenger et al. 2003a; Floss et al. 2006; Busemann et al. 2009a). Other presolar grains typically found in meteorites in lower abundances, such as SiC, graphite, and oxides (e.g., Davidson 2009; Davidson et al. 2009, Forthcoming; Floss and Stadermann 2009a; Nguyen and Messenger 2011; Leitner et al. 2012a), are rarely seen in IDPs. Exceptions include several C-rich grains with isotopic anomalies (Floss et al. 2004; Busemann et al. 2009a), one aluminum oxide (Stadermann et al. 2006), and two SiC grains (one confirmed and one described as potential; Stadermann et al. [2006] and Busemann et al. [2009a], respectively).

Two groups of isotopes (see the Experimental section) were measured in order to locate refractory C-rich phases (including presolar SiC and other isotopically anomalous C grains) and O-anomalous grains (presolar oxides and silicates). A total of 14 presolar grains were located in 10 IDPs; one SiC, one other isotopically anomalous C grain (possibly graphite), ten silicates and two oxides (Table 2). A number of D- and 15N-hotspots, possibly associated with organic matter, were also found (Table 2).

The majority of anomalies (13 of 14 presolar grains; 35 of 38 other isotopic anomalies) are located in non-GSC IDPs (Figs. 2 and 3); the total area ratio of the two groups (non-GSC:GSC IDPs) is only 3:2. IDP Pizarro (which fragmented into four pieces, a–d, during picking; Fig. 2) shows the most variation in the anomalies present: one presolar SiC grain, two silicates, and one oxide, as well as N isotopic anomalies (Fig. 2), attesting to its unaltered nature. The fact that it fragmented during picking also hints to its primitive nature as more primitive IDPs are friable and often fragment on impact with collector flags producing cluster IDPs (Messenger 2000; Messenger et al. 2003b; Busemann et al. 2009a). IDPs Magellan, Polo, Hudson, Raleigh (non-GSC IDPs; Fig. 3), and Drake (GSC IDP) are also rich in N and D isotopic anomalies suggesting that they have remained largely unaltered since their formation and thus are relatively pristine.

Figure 2.

 Secondary electron images of IDP Pizarro fragments (a–d), with annotations showing locations of all isotopically anomalous areas (outlines), presolar grains (arrows and circles). Numbers correspond to those in Tables 3–5. Extremely 17O-rich presolar silicate “b Sil50” is marked (see Fig. 4a for O-isotopic composition). All images are at the same scale.

Figure 3.

 (a) Secondary electron image of non-GSC IDP Raleigh, and (b–d) NanoSIMS isotopic ratio maps showing: (b) three H isotopic anomalies (or “hotspots”), (c) one 17O isotopic anomaly identified as a presolar silicate (the O-anomalies spatially correlate with 28Si), and (d) three N isotopic anomalies. Variations in isotopic ratios are given as deviations from terrestrial values in permil (δ notation), and are corrected for instrumental mass fractionation. Labels correspond to those in Tables 3–5. All scale bars are 1 μm.

C-Anomalous Phases: Presolar SiC and Other

Phases anomalous in C isotopes can be broadly sub-divided into two categories: presolar SiC and non-SiC (presolar graphite, unspecified organic matter, and nanoglobules). Isotopically anomalous SiC can be distinguished from other isotopically anomalous C-rich grains, which do not contain significant Si, on the basis of their Si/C ratios. Non-SiC isotopically anomalous carbons, with the exception of presolar graphite grains, i.e., OM and nanoglobules, are often associated with D-enrichments. The apparent lack of abundant grains with significant C isotopic anomalies (particularly those identified as presolar SiC) in IDPs may be somewhat surprising (Floss et al. 2004, 2006). Only one confirmed presolar SiC grain (Stadermann et al. 2006) and few other C-anomalous grains (Floss et al. 2004, 2010) have been reported in IDPs. However, given the small sizes of IDPs the lack of abundant C isotopic anomalies (SiC and other) is perhaps not so surprising. For instance, if a similar presolar SiC abundance as that found in the primitive CR chondrites (35 ± 4 ppm average; Davidson et al. forthcoming) is assumed for IDPs, and the average pressed IDP is 10 μm diameter (with a circular surface area of 80 μm2), a total of 30 IDPs would have to be analyzed in order to find a single presolar SiC grain 350 nm in diameter (with circular surface area of 0.1 μm2).

Carbon isotopic data for all IDPs studied here revealed four C anomalies; one is spatially correlated with Si, and is thus a presolar SiC grain (SiC9 in Pizarro a; Table 3; Fig. 2), another is a submicron region slightly enriched in 12C with isotopically normal N (C3 in Hudson; Table 3), which is not spatially correlated with Si. The remaining two (Columbus N2 and Pizarro N4) C isotope anomalies are associated with isotopically anomalous N. The discovery of four more isotopically anomalous C-grains is significant as it increases the number found to date in IDPs considerably. The bulk δ13C values of the IDPs analyzed here range from 7‰ to 23‰, which lie within the typical range observed previously in IDPs (e.g., Messenger et al. 2003b; Floss et al. 2006; Busemann et al. 2009a). The presolar SiC grain (approximately 200 nm in diameter) identified in “Pizarro a” is enriched in 12C (12C/13C = 240 ± 60, δ13C = −620 ± 100 ‰) relative to solar composition (12C/13C = 89), has N (14N/15N = 380 ± 120, δ15N = −280 ± 230‰) of solar composition (14N/15N = 272) within error, δ29Si = 25 ± 12‰, and δ30Si = 21 ± 16‰; suggesting it is a Y grain and thus formed in the outflows of an asymptotic giant branch (AGB) star (Amari et al. 2001).

The non-SiC C-anomalous phase in Hudson has C (12C/13C = 112 ± 6, δ13C = −210 ± 40‰) and N (14N/15N = 290 ± 25, δ15N = −60 ± 80‰) isotopic compositions that fall within the expected range for presolar graphite (Nittler 2003). However, without structural analysis of the grain it was not possible to determine its specific nature (i.e., presolar graphite or nanoglobules/monolithic amorphous C; Busemann et al. 2006b; Nakamura-Messenger et al. 2006) as it was completely sputtered away during subsequent analyses. However, this grain had a δ13C of approximately −200‰, which is much closer to the ±100‰ range reported in OM (Floss et al. 2004, 2006; Busemann et al. 2006b, 2009a; Nakamura-Messenger et al. 2006) than typical anomalies in presolar graphite grains (Amari et al. 1990). This C-anomaly was not correlated with any other isotopic anomalies in H, N, O, or Si. A similar 13C-depleted anomaly from an IDP was reported by Floss et al. (2004). However, this phase was also isotopically anomalous in its N (12C/13C = 96.6 ± 1.3, δ13C = −70±13‰; 14N/15N = 119.8 ± 1.3, δ15N = 1270 ± 25‰) and much larger in size (0.6 × 1.8 μm compared to approximately 0.2 × 0.2 μm seen here). Similar C-anomalies were also found in the CR carbonaceous chondrites EET 92042 (as a micron-size grain; δ13C = −113‰, 12C/13C = 100; δ15N = 1150‰, 14N/15N = 127; Busemann et al. 2006a), and QUE 99177 (18 grains; Floss and Stadermann 2009b) and MET 00426 (10 grains; Floss and Stadermann 2009b).

O-Anomalous Phases: Presolar Silicates and Oxides

NanoSIMS 17O/16O and 18O/16O ratio maps revealed the presence of 12 O-anomalous presolar grains in 7 of the 10 IDPs analyzed (Table 4). Eleven of these grains (nine silicates, total abundance 2800 ± 900 ppm; and two oxides, 800 ± 500 ppm) were located in the non-GSC IDPs; only one grain (a silicate, 920 ± 700 ppm) was located in the GSC IDPs (Table 4). This is not surprising given the hydrated nature of the other two GSC IDPs analyzed (Table 1) as, e.g., Floss and Stadermann (2005) suggested that presolar silicates are progressively destroyed by increasing aqueous alteration. The O isotopic compositions of all presolar silicate and oxide grains are within the ranges previously reported for those in other IDPs and meteorites (Fig. 4; Nittler et al. 1997; Messenger et al. 2003a; Nittler 2003; Nguyen et al. 2007b; Vollmer et al. 2008; Busemann et al. 2009a). However, one silicate (Sil50 from Pizarro b) appears to be the most 17O-rich silicate reported in an IDP to date (Fig. 4a) and one of the most 17O-rich silicates reported in any extraterrestrial material (e.g., Nguyen et al. 2007b, 2010; Hynes and Gyngard 2009; Bose 2011). Interestingly, this grain is, as identified by FIB/TEM analysis (Davidson 2009), a possible presolar GEMS aggregate, further attesting to the anhydrous pristine nature of this IDP. Unfortunately, the individual components of the “GEMS” were too small to obtain SAED patterns from, and the glassy areas suffered significant beam damage when EDS X-ray spectra were collected (Davidson 2009). However, EDS on small constituent grains showed them to be very Fe-rich with some Ni and S, indicating that they are metals and sulfides. Presolar GEMS grains have been previously identified, though rarely, in both IDPs (e.g., Messenger et al. 2003a; Keller and Messenger 2011) and chondritic meteorites (e.g., Mostefaoui et al. 2004; Nguyen et al. 2010).

Figure 4.

 Oxygen isotopic compositions of 12 anomalous-oxygen grains, (a) ten presolar silicates, and (b) two presolar oxides. Also shown are pre-existing O-anomalous grains found in IDPs (grey circles) from Messenger et al. (2003a), Floss et al. (2006, 2010), Stadermann et al. (2006), and Busemann et al. (2009a). Numbers in (b) correspond to the presolar oxide groups identified by Nittler et al. (1997). All presolar silicates found here and one oxide plot in Group 1 (the most common group). The remaining oxide appears to be from Group 4. Dashed lines, marked as “solar” indicate the solar ratios for 17O/16O and 18O/16O. 1σ errors are smaller than the symbols.

The mineralogy (in terms of oxide versus silicate) of each O-anomalous presolar grain was roughly determined on the basis of their 28Si/16O and (mass-40)/16O ratios (Zinner et al. 2003; Nguyen et al. 2007b). Silicate grains have 28Si/16O ratios similar to bulk IDP 28Si/16O (IDPs are primarily composed of silicates of solar composition). In this instance, presolar silicates have 28Si/16O ratios of 0.02–0.04, whilst presolar oxides are identified by 28Si/16O ratios of 0.01 (Table 4).

Presolar silicate abundances found in this work are much higher for the non-GSC IDPs (2800 ± 900 ppm) than the anhydrous GSC IDP Drake (920 ± 700 ppm) despite large errors due to poor statistics, and significantly higher than previous data (Fig. 5) with the exception of two GSC IDPs reported by Busemann et al. (2009a), with abundances of 15,000 and 3500 ppm. These GSC IDPs (Busemann et al. 2009a) have the highest presolar silicate abundances reported in extraterrestrial materials to date, attesting to the highly primitive nature of these IDPs. The highest abundance seen here, albeit with a large error, was 6300 ± 4000 ppm for Pizarro; much higher than previous reports for IDPs of 120–375 ppm (Floss et al. 2006), and the most primitive meteorites (typically 100–200 ppm for the most primitive carbonaceous chondrites Acfer 094, ALHA 77307, MET 00426, and QUE 99177; Nguyen et al. 2007b, 2010; Vollmer et al. 2008; Floss and Stadermann 2009).

Figure 5.

 The presolar silicate grain abundances for IDPs studied here (grey bars) compared to previously reported abundances for other IDPs, primitive meteorites (matrix-normalized) and stardust samples from comet 81P/Wild 2 (black bars). A presolar silicate abundance of 500 ppm is shown for all GSC IDPs reported by Busemann et al. (2009a), and 3500 ppm and 15000 ppm for their GSC IDPs E1 and G4, respectively. One IDP (Andric) from the 55P/Tempel-Tuttle comet collection (Floss et al. 2010) is also compared. References: aBusemann et al. (2009a), bFloss et al. (2010), cFloss et al. (2006), dFloss and Stadermann (2008), eNguyen et al. (2007a), fVollmer et al. (2008), gStadermann and Floss (2008), and hLeitner et al. (2012b). *Stadermann and Floss (2008) and Leitner et al. (2012b) calculated bulk presolar O-anomalous grain abundances for 81P/Wild 2 samples; no compositional information (i.e., silicate versus oxide) is available for O-anomalous presolar grains from these samples. The abundance of approximately 1100 ppm is based on the detection of one presolar grain and hence still associated with a large uncertainty (Leitner et al. 2012b).

Isotopically Anomalous Organic Matter

IDPs generally exhibit abundant and large enrichments in D and 15N relative to terrestrial isotopic ratios (Messenger 2000; Messenger et al. 2003b), probably mostly carried by OM that is enriched in D and 15N (Busemann et al. 2009a), which attests to the primitive nature of the IDPs. The D/H ratios of the cluster IDPs studied here are generally slightly higher than those previously reported for non-GSC cluster IDPs (Fig. 6; Floss et al. 2006). However, the D-hotpots within the IDPs studied here have D/H ratios lower than the bulk composition of one GSC IDP reported by Busemann et al. (2009a). Of the three largest bulk IDP D-enrichments (Columbus, δD = 1525 ± 80‰; Raleigh, δD = 1780 ± 60‰; and Drake, δD = 1140 ± 60‰) two are non-GSC IDPs (Columbus and Raleigh). No “hotspots” were identified in GSC IDPs studied here. Non-GSC IDPs Magellan, Columbus, and Raleigh each contain one D-hotspot 1 μm in diameter with δD of 3500–3700‰ (Table 5), the largest D enrichments found in this study. No extreme D-enrichments like those reported in previous studies of GSC IDPs (δD up to 29,000‰; Busemann et al. 2009a) were seen.

Figure 6.

 Ranges of hydrogen isotopic compositions (D/H ratios) in IDPs studied here and in the literature, normalized to SMOW (standard mean ocean water). Lines indicate the range of D/H ratios seen in each IDP fragment. Each point represents a different measurement within the same IDP. Circles are bulk values, triangles are “coldspot,” and squares are “hotspot” anomalies within that IDP, shown where possible for IDPs reported here and from Floss et al. (2006; Cameca ims-3f ion microprobe) and Busemann et al. (2009a; Cameca ims-6f ion microprobe). The shaded region shows the range of terrestrial D/H values (Floss et al. 2006). After Fig. 2 in Floss et al. (2006).

The 15N/14N bulk ratios of all IDPs studied here are similar to those seen in other IDPs (Messenger 2000; Aléon et al. 2003; Messenger et al. 2003b; Floss et al. 2004, 2006; Busemann et al. 2009a). Bulk 15N/14N values of the GSC IDPs are in agreement with the non-GSC IDPs. Whilst 15N-hotspots were seen in all seven non-GSC IDPs, only one of the three, GSC IDP (Drake), was found to contain any. However, the latter were very large and cover twice the relative area (in percentage of area analyzed) of all those in non-GSC IDPs (Table 7). The most extreme 15N-hotspots were found in Cortes, Magellan, and Columbus; the latter two of which also contain the two largest D-enrichments (Tables 1 and 5). Although there is some overlap between regions enriched in 15N and D (e.g., Raleigh; Figs. 3d and 3h) this was not commonly seen.

Raman Spectroscopy

Raman spectroscopy has previously been used to investigate the relative degree of disorder/maturity of organic matter in IDPs (e.g., Wopenka 1988; Raynal et al. 2001; Raynal 2003; Quirico et al. 2005; Muñoz-Caro et al. 2006; Nittler et al. 2006; Stadermann et al. 2006; Rotundi et al. 2007; Busemann et al. 2009), Stardust samples from comet 81P/Wild 2 (Sandford et al. 2006; Rotundi et al. 2008), and meteorites (e.g., Raynal et al. 2001; Quirico et al. 2003, 2005; Raynal 2003; Bonal et al. 2006; Busemann et al. 2007). Raman spectra of samples that contain disordered carbonaceous material are dominated by two bands: the D (“disordered”) and G (“graphite”) bands centered near 1360 and 1590 cm−1, respectively (Rotundi et al. [2008] and references therein). The highly ordered nature of pure graphite yields a single, narrow G band around 1582 cm−1. With increasing disorder, both bands broaden significantly, other bands (such as the so-called D′ at approximately 1620 cm−1) may begin to appear, and the apparent G band position moves towards higher wavenumbers (Fig. 7). The D and G band peak parameters relative intensities (I), full widths at half maximum (Г), and positions (ω) detected in the OM are correlated with the alteration processes it experienced. Alteration processes may include amorphization resulting from irradiation in space, thermal metamorphism on parent bodies, or changes in chemical composition (e.g., Busemann et al. 2007).

Figure 7.

 Raman (a, c, e) D- and (b, d, f) G-band parameters (ω = line position; Γ = full width at half maximum) of (a, b) the IDPs studied here, compared with those of (c, d) other IDPs and Stardust, and (e, f) meteoritic insoluble organic matter. References: 1GSC IDPs E1 and G4; Busemann et al. (2009a), 2IDPs; Muñoz-Caro et al. (2006), Stadermann et al. (2006; G-band only), Rotundi et al. (2007), 3IDPs; Raynal (2003), 4Stardust; Rotundi et al. (2008), 5IOM; Busemann et al. (2007). All data were collected with 514.5 nm lasers with the exception of data for GSC IDPs E1 and G4 (Busemann et al. 2009; 532 nm) and data from Stadermann et al. (2006; 532 nm). Downshifts of 4 Δcm−1 were applied to ωD and ωG data from Busemann et al. (2009a) to allow for the comparison between their data measured with 532 nm laser light compared to 514.5 nm at the OU here (see Rotundi et al. 2008). Dashed lines and arrows indicate the increasing order of primitiveness (Busemann et al. 2007). After Fig. 2 in Rotundi et al. (2008).

The organic matter in the IDPs studied here is disordered and hence primitive in the sense that it has not been significantly heated on the parent bodies or during atmospheric entry (Fig. 7). It is comparable to previous studies of IDPs (e.g., Muñoz-Caro et al. 2006; Stadermann et al. 2006; Rotundi et al. 2007), and Stardust samples from comet Wild 2 (Rotundi et al. 2008), and is slightly more disordered than meteoritic IOM (Figs. 7e and 7f) (Busemann et al. 2007). The organic matter in the anhydrous GSC IDP Drake (Fig. 7) is the most disordered among the organic matter of all IDPs studied here and is more disordered than many of those reported in the literature. Hawkins and Frobisher did not yield useful D and G bands sufficient for fitting. On the basis of D-band parameters, non-GSC IDP Raleigh also appears to be very primitive (Fig. 7). Relative to the extremely primitive GSC IDP Drake, the less primitive nature of the other non-GSC IDPs is consistent with their lower bulk H isotopic compositions (Table 5; Fig. 6), with the exception of Columbus, which has the second largest D-enriched bulk composition (δD = 1500‰) of all the IDPs studied here but appears to be one of the more processed IDPs in terms of Raman analyses (Figs. 7a and 7b).


Defining the Primitive Nature of IDPs: Anomalous Bulk Nitrogen

An isotopically primitive subgroup of IDPs were distinguished by Floss et al. (2006) on the basis of their anomalous bulk N compositions. Similarly, Busemann et al. (2009) found that the most primitive GSC IDPs (based on presolar grain abundances and Raman spectral parameters) also have 15N-enriched bulk compositions. However, no correlation between “primitiveness” (based on the disorder of organic matter, bulk isotopic compositions, and abundances of presolar/other isotope anomalies) and bulk N was seen here (Fig. 8). The isotopically primitive subgroup identified by Floss et al. (2006) consisted of IDPs with anomalous bulk N, which contained all presolar silicates and the majority of 15N-hotspots identified in that study. Here, abundant isotopic anomalies and presolar grains were found in IDPs that have normal bulk N isotopic compositions (Table 3). The three GSC IDPs analyzed here (including two hydrated ones) have no, or few, hotspot anomalies but still have varying degrees of anomalous bulk N isotopic compositions (Table 3). This evidence weakens the classification used by Floss et al. (2006), who suggested that identifying IDPs with bulk anomalous N will allow for very primitive presolar materials to be more easily identified. Whilst this may be true in most cases, the use of such an approach may bias the unambiguous identification of primitive IDPs, as those with isotopically normal bulk N studied here also contain abundant presolar grains and other isotopic anomalies and, hence, are primitive IDPs too. The differences in bulk N and abundance of presolar components seen between the IDPs studied here and by Floss et al. (2006) may have resulted from different groups of IDPs sampling different regions of the early solar system (none of the IDPs in this study were from the same collector flags as those in the Floss et al. study) or, more likely, the difference results from heterogeneity of the parent body/bodies at the 10–100 μm scale (hinted at by the 250‰ variation observed in Pizarro, Cortes, and Magellan, which are all fragments from the same cluster particle; Table 3).

Figure 8.

 Comparison of bulk isotopic compositions and Raman spectral parameters from GSC and non-GSC IDPs, including bulk N-isotopic composition (δ15N in per mil) versus (a) D-band position, (c) G-band position, and (e) bulk C-isotopic composition (δ13C in per mil), and bulk H-isotopic composition (δD in per mil) versus (b) D-band position, (d) G-band position, and (f) bulk N-isotopic composition (δ15N in per mil).

Grigg-Skjellerup Collection versus Non-GSC IDPs

Evidence for the primitive nature of the IDPs studied here includes anomalous bulk and hotspot/coldspot isotopic compositions (Figs. 4 and 6, Tables 3–5), disordered character of organic matter (Fig. 7; Table 6), presolar grain abundances and distribution (Fig. 5; Table 2), and IDP textures (Fig. 1). These primitive characteristics have been previously reported in other IDPs (e.g., Messenger et al. 2003a, 2005; Floss et al. 2006, 2010; Busemann et al. 2009a), and their abundant presence in some of the IDPs analyzed here affirms their highly primitive nature. In order to determine their relative degrees of primitiveness, and identify any differences between GSC and non-GSC IDP populations, the different parameters listed above have been compared (Table 7).

Table 6. IDPs and their Raman spectral parameters (band intensity I, position ω, and full width at half maximum Г) for the D (“disordered”) and G (“graphite”) bands.
NameTotal spectraωD (cm−1)aГD (cm−1)aωG (cm−1)aГG (cm−1)aID/IGa
  1. aUncertainties are 1σ standard deviation of the mean.

  2. bNo useful spectra.

  3. cMeteoritic IOM data from this study agree well with previous data (Busemann et al. 2007) confirming the validity of IDP data presented here and comparison to literature data (e.g., Rotundi et al. 2008; Busemann et al. 2009).

Non-GSC cluster IDPs
Pizarro191365.6 ± 3.5349.6 ± 24.41577.7 ± 4.2132.2 ± 13.01.162 ± 0.113
Cortes61369.6 ± 4.2331.3 ± 45.21579.0 ± 1.7116.8 ± 12.81.130 ± 0.095
Magellan91367.7 ± 3.2338.4 ± 12.11582.8 ± 2.2131.0 ± 16.31.105 ± 0.031
Columbus51361.6 ± 2.0287.4 ± 15.81584.7 ± 2.2104.5 ± 3.31.218 ± 0.030
Polo21363.3 ± 3.7284.1 ± 0.51582.1 ± 2.9101.6 ± 2.61.180 ± 0.072
Hudson81362.8 ± 3.7305.0 ± 23.41582.6 ± 2.5108.5 ± 6.81.071 ± 0.045
Raleigh41370.8 ± 6.7411.7 ± 40.61581.2 ± 0.9114.4 ± 7.11.330 ± 0.141
GSC cluster IDPs
Drake151375.8 ± 4.3346.7 ± 49.51567.7 ± 3.1115.4 ± 3.31.155 ± 0.163
Primitive carbonaceous chondrite insoluble organic matter (IOM)c
Murchison131357.7 ± 3.5316.2 ± 17.61592.2 ± 0.8107.1 ± 8.21.059 ± 0.036
Cold Bokkeveld161358.9 ± 2.0291.5 ± 11.81590.7 ± 2.496.8 ± 4.60.988 ± 0.015
Leoville371352.5 ± 3.3231.5 ± 9.61588.1 ± 3.480.1 ± 2.41.069 ± 0.035
Allende161347.0 ± 1.679.1 ± 2.81591.6 ± 2.464.6 ± 1.21.544 ± 0.073
Table 7. Comparison of different parameters for non-GSC IDPs and one GSC IDP (Drake).
ParametersNon-GSC IDPsaGSC IDP (Drake)a
  1. aErrors are 1σ.

  2. bBased on a single data point.

Total number IDPs71
Total IDP area (μm2)392173
Presolar grain abundances (number/ppm)
 SiC1/70 ± 600/0
 C-anom.1/100 ± 800/0
 Silicates9/2800 ± 9001/920 ± 700
 Oxides2/800 ± 5000/0
Bulk isotopic compositions
 δ15N (Min)−167 ± 8239 ± 2
 δ15N (Max)243 ± 9
 δ13C (Min)7 ± 7−1 ± 2
 δ13C (Max)23 ± 4.9
 δDSMOW (Min)297 ± 461141 ± 59
 δDSMOW (Max)1787 ± 62
Isotopic anomalies
 δ15N (Min)−281 ± 231633 ± 58
 δ15N (Max)1078 ± 89835 ± 70
 δDSMOW (Min)−698 ± 60
 δDSMOW (Max)3718 ± 1197
 δ17O (Max)15715 ± 435270 ± 49b
 δ18O (Max)107 ± 8−47 ± 18b
Isotopic anomalies (number/area%)

GSC versus Non-GSC IDPs: Presolar Grain Abundances

A variety of presolar grain types were found in the IDPs studied here; the most abundant of which were presolar silicates. Despite their relative scarcity (by number) in the IDP record, one presolar SiC, two nonspecific C-rich grains, and two presolar oxide grains were located in two of the seven non-GSC IDPs studied here (Pizarro and Cortes; from the same cluster). This yields presolar grain abundances that are extremely high compared to almost all previous IDP sets (e.g., Floss et al. 2006, 2010; Busemann et al. 2009a) and all meteorite samples (e.g., Floss and Stadermann 2009; Nguyen and Messenger 2011; Davidson 2009; Davidson et al. forthcoming; Leitner et al. 2012a; Floss and Stadermann 2012).

The Non-GSC IDPs of this study yield a total presolar silicate abundance of 2800 ± 900 ppm; the only presolar silicate from the anhydrous GSC IDP Drake yields an abundance of 920 ± 700 ppm. While these abundances are lower than those reported for particularly presolar silicate-rich individual IDPs (5500 ppm; Messenger et al. 2003b), and two from the comet Grigg-Skjellerup collection (15000 and 3500 ppm; Busemann et al. 2009a), they are much higher than those abundances reported by Floss et al. (2006), including for their isotopically primitive, 15N-rich subgroup. They are also higher than the overall abundance of 500 ppm given for all GSC IDPs studied by Busemann et al. (2009a), and for one IDP from the comet Tempel-Tuttle targeted collector (700 ± 300 ppm; Floss et al. 2010). While a particularly large silicate grain, approximately 1 μm in diameter, may have skewed the Messenger et al. (2003b) abundance, there is no evidence to suggest that the presence of anomalously sized grains increased the abundances reported here. The abundance of presolar silicates seen here in the GSC IDPs may seem lower than expected when compared with the high abundances of GSC IDPs reported by Busemann et al. (2009a). However, Busemann et al. (2009a) reported the study of five cluster and three individual GSC IDPs and only two contained O-anomalous grains (presolar silicates were very abundant in those IDPs; four and seven grains). Similarly, Floss et al. (2010) analyzed microtomed slices of two cluster IDPs (total area = 330 μm2) from comet Grigg-Skjellerup targeted collectors and neither contained any presolar grains. Messenger (2002) predicted that 1 to 50% of the IDP flux collected after the passing of comet Grigg-Skjellerup would be from the comet itself. Therefore, it is not surprising that only one presolar silicate was detected here as only three IDPs of potential comet Grigg-Skjellerup origin were analyzed. Material returned from comet Wild 2 yielded very low bulk presolar O-anomalous (i.e., presolar silicate/oxide) abundances of 10–20 ppm; only four presolar silicate or oxide grains (McKeegan et al. 2006; Stadermann and Floss 2008; Stadermann et al. 2008; Leitner et al. 2010) and one presolar SiC grain (Brownlee et al. 2009; Messenger et al. 2009) have been discovered. However, this may be the result of alteration during capture or dispersion of the fine-grained original Wild 2 material in the aerogel along the complete track that cannot be easily scanned for presolar grains, and therefore not a true reflection of the comet’s presolar grain inventory (Stadermann et al. 2009). Leitner et al. (2012b) estimated a much higher O-anomalous grain abundance (1100 ppm) for Wild 2 material, albeit with large and currently unspecified uncertainty, as this number is based on one presolar grain.

The presence of clusters of presolar grains in some IDPs, such as those in Hudson and Pizarro, and not others implies that these IDPs inherited a heterogeneous mix of presolar grains and primitive materials from the early solar system. “Clusters” of primitive materials in extraterrestrial materials have been observed previously (e.g., Yada et al. 2008; Busemann et al. 2009a; Floss and Stadermann 2009a; Vollmer et al. 2011). By this logic, it is possible that some primitive IDPs may inherently lack presolar grains, and thus have low grain abundances, at a 10–100 μm scale, which may not necessarily result from destruction of grains by alteration, making it difficult to use presolar grain abundances in a few IDPs alone as a measure of primitiveness. However, in the case of Hawkins and Frobisher (which have been hydrously altered) presolar silicates may have simply been destroyed (e.g., Floss and Stadermann 2005; Trigo-Rodriguez and Blum 2009; Leitner et al. 2012a).

It should be stressed that the total surface area of IDPs studied here (Table 7) and elsewhere is very small and thus analyses are based on limited statistics (as reflected by large estimated abundance errors). In the case of presolar SiC, isotopically anomalous carbon, and oxides only one or two grains, were found of each type in non-GSC IDPs compared to no grains of these type in GSC IDPs. As a result, based on this study there appears to be no statistically significant difference between non-GSC IDPs and GSC IDPs (e.g., Busemann et al. 2009), in terms of presolar grain abundances. Therefore, the detection of potential differences between these two (and any other “targeted collection”) IDP populations requires a larger number of IDPs for statistically robust results. On the other hand the actual presence of a large number of presolar grains (and other pristine matter) in a single IDP indeed hints at the little alteration of this IDP in space and an origin in either a comet or a primitive outer main-belt asteroid not sampled by any meteorites to date.

GSC versus Non-GSC IDPs: Raman Spectroscopy

It is difficult to compare the GSC and non-GSC IDPs on the basis of Raman parameters in this study since data only exist for one GSC IDP (Drake). Nevertheless, on the basis of Raman spectral parameters alone GSC IDP Drake appears to be more primitive than the non-GSC IDPs, and many others in the literature (see Figs. 7c and 7d). Whilst there appears to be no correlation between bulk D/H and G-band position for OM (Fig. 8d) within the non-GSC IDPs, the GSC IDP Drake is quite distinct from the non-GSC IDPs on the basis of these parameters and clearly contains the most primitive OM detected in this study. Busemann et al. (2009b) compared Raman spectral parameters for 21 GSC IDPs with 15 IDPs collected at different times (i.e., no potential known source), and found that the variation among GSC IDPs is larger than that of the non-GSC IDPs, with the most primitive IDPs being mostly GSC IDPs. However, both sets of IDPs show similar average values. The Raman results obtained in this study (Figs. 7 and 8) appear to reinforce the distinction between Raman parameter variability obtained from GSC and non-GSC IDPs stated by Busemann et al. (2009).

Non-GSC IDPs with highly disordered OM include Pizarro and Hudson (Figs. 7a and 7b). Pizarro is associated with IDPs Cortes and Magellan (all from collector L2005, cluster #31; Table 1), and Hudson is associated with Raleigh (collector L2006, cluster #3; Table 1). These IDPs are also primitive in nature but do not show such elevated presolar grain abundances as those seen in Pizarro and Hudson, once again demonstrating the heterogeneity of IDPs at the 10–100 μm scale.

GSC versus Non-GSC IDPs: Isotopic Anomalies

Anomalous H isotopic compositions were first reported in IDPs by Zinner et al. (1983), indicating their extraterrestrial and extraordinarily primitive nature. Since high D/H ratios have been observed in simple molecules (e.g., H3+, HD, H2D+, CH3+) in cold molecular clouds (e.g., Millar et al. 1989), Messenger (2000) suggested that this indicates the survival of interstellar molecules in some IDPs. Deuterium anomalies (enrichments and depletions) have since been reported in many IDPs, with hotspots (i.e., enrichments) of up to δDSMOW = 50,000‰ (Messenger et al. 1995). All IDPs studied here show similar ranges of bulk D-isotopic compositions, δD = 300 to 1800‰, with hotspots of 2200 to 3700‰. The non-GSC IDPs Raleigh and Columbus have the most anomalous bulk values. While GSC IDP Drake is particularly enriched in deuterium, the bulk δD-compositions of the other GSC IDPs (hydrated Hawkins and Frobisher) are some of the closest to terrestrial of all the IDPs studied here. Although bulk values seen here are generally higher than the bulk values reported by Floss et al. (2006), no extreme localized enrichments like those reported by Messenger (2000) and Busemann et al. (2009a) for IDPs or in Antarctic micrometeorites (Duprat et al. 2010) were seen. The spatial resolution achieved during the analyses reported here is comparable to, or better than, previous studies and therefore it is unlikely that the lack of such anomalous regions reported here is due to the analytical method. Although the majority of D-anomalies reported to date are enrichments a few depletions have also been seen. However, these are relatively scarce, because it is more difficult to obtain statistically significant data for depletions in low abundance isotopes (Floss et al. 2006). Whilst D hotspot anomalies were seen in the majority of non-GSC IDPs none were found in any of the GSC IDPs (Tables 5 and 7).

Comparison of C and N maps with D-anomalies shows that one large D-hotspot appears to overlap slightly with a 15N-hotspot (Figs. 3b and 3d). However, this is unlikely to be a true reflection of the composition of the D-hotspot, since significant material was sputtered away between the N and H measurements when O-isotopes were analyzed to locate presolar silicates and oxides. More likely the anomalous N-bearing phase was completely sputtered away and is not associated with the D-hotspot supporting the view that the carriers of 15N and D are mostly distinct (e.g., Keller et al. 2004; Busemann et al. 2006b, 2009a), although possibly not always (e.g., Aléon 2010). The observation of generally decoupled D and 15N enrichments appears in agreement with most recent calculations and astronomical observations (Wirström et al. 2012).

Although not as extreme as D-anomalies, hotspot 15N enrichments are commonly seen in IDPs (e.g., Stadermann et al. 1989; Stadermann 1991; Messenger et al. 2003a; Floss et al. 2006, 2010; Busemann et al. 2009a). All non-GSC IDPs display localized 15N-enrichments (Tables 3 and 7), whereas their bulk N isotopic compositions are only slightly anomalous, with most being close to solar composition. Of the three GSC IDPs all have slightly anomalous bulk N isotopic compositions but only the anhydrous IDP Drake contains 15N-hotspots. These three hotspots are very large and cover approximately 1% of the surface area of all the GSC IDPs, compared to approximately 1.2% for 28 smaller anomalies in all non-GSC IDPs (Table 7). However, all GSC and non-GSC IDPs cover a similar range of bulk N isotopic compositions. The range of hotspot anomalies seen in the IDPs studied here (δ15N = 122 to 1100‰) is consistent with those reported by others (e.g., Mukhopadhyay et al. 2002; Floss et al. 2006, 2010; Busemann et al. 2009a), although this range does not reach the most 15N-enriched hotspot reported in the GSC IDPs reported by Busemann et al. (2009a; δ15N = 1470 ± 220‰).

IDP Hawkins shows a significant (i.e., >3σ error) bulk depletion in 15N (δ15N = −190 ± 27‰). The only other IDPs reported to date with bulk 15N-depletions are Porky (δ15N = −93 ± 4‰; Messenger 2000) and Eliot (δ15N = −108 ± 9‰; Floss et al. 2006). Messenger (2000) suggested a nucleosynthetic origin for the 15N depletion in Porky. However, such N isotopic compositions fall between solar, with δ15N almost −400‰ (Marty et al. 2011), and the 15N enriched compositions observed in many IDPs, and therefore may reflect simply a more or less extensive exchange between solar gas and the organics, depending on the starting 15N-enrichment of the organic matter. Interestingly, N isotopic compositions of HCN (which is considered to be a primary component of cometary ices) in comet Hale-Bopp, δ15N = −157 ± 120‰ (Jewitt et al. 1997) and δ15N = −174 ± 245‰ (Ziurys et al. 1999), agree with Eliot (Floss et al. 2006) and Hawkins measured here. However, more recent analyses (Bockelée-Morvan et al. 2008) indicate that N in comets is generally enriched in 15N.

Whilst the GSC IDPs analyzed here lack abundant hotspot N, C, and H isotopic anomalies, their bulk N, C, and H isotopic compositions are in agreement with the non-GSC IDPs, thus the two groups of IDPs analyzed here cannot be distinguished on the basis of these parameters.

Cometary Origins of IDPs

Until recently, hydrous IDPs were generally thought to be of asteroidal origin (e.g., Bradley and Brownlee 1986; Keller et al. 1992; Rietmeijer 1992), whilst fine-grained anhydrous IDPs were thought to be cometary (e.g., Brownlee et al. 1995; Bradley 2003). The following observations are consistent with the view that all non-GSC IDPs from this study and GSC IDP Drake (i.e. all CP-IDPs) are of cometary origin: fine-grained porous textures, anhydrous nature of their mineral components, and an abundance of easily-destroyed presolar silicates (e.g., Brownlee et al. 1995; Bradley 2003). Also, the isotopic compositions of many N-hotpots seen in all anhydrous IDPs studied here (Table 3) are similar to N isotopic compositions derived from CN spectra of 23 comets with 14N/15N of 147 ± 5 (Manfroid et al. 2009). However, relatively recent astronomical observations of comet 9P/Tempel 1 during the Deep Impact encounter found evidence for the presence of amorphous silicates and carbon, carbonates, phyllosilicates, polycyclic aromatic hydrocarbons, and water gas and ice (Lisse et al. 2006), indicating that cometary IDPs may also contain hydrous mineral phases (see also Busemann et al. 2009a). Stardust samples (from comet 81P/Wild 2) were initially thought to be more similar to anhydrous IDPs than to any other meteoritic material (Zolensky et al. 2006) but it now seems that comet 81P/Wild 2 has more similarities with chondritic asteroids (Ishii et al. 2008), particularly since it contains high-temperature phases that may have formed in the inner solar system (e.g., Nakamura et al. 2008). Furthermore, the discovery of main belt comets, evidence for water ice on the surface of asteroid 24 Themis (Hsieh and Jewitt 2006; Hsieh et al. 2009; Campins et al. 2010; Rivkin and Emery 2010), recent oxygen isotope studies (which show similar compositions for carbonaceous chondrites, IDPs, and comet Wild 2; Nakashima et al. [2012]; and references therein), and the identification of aqueously altered material in Stardust samples from Wild 2 (Berger et al. 2011) has blurred the traditional boundary between comets and asteroids, suggesting that water played an important role in the formation and evolution of minor bodies in the solar system.

Grigg-Skjellerup Collection IDPs and Comet Grigg-Skjellerup

At present, it is difficult to compare collections of targeted and nontargeted IDPs because of low statistics (see above and e.g., Floss et al. 2010); IDPs are inherently small in size and number when compared with other extraterrestrial materials and typical studies rarely extend beyond the analysis of 10 different cluster particles (e.g., Floss et al. 2006, 2010; Busemann et al. 2009a; this study). The situation is further complicated because, as shown here by non-GSC IDP Pizarro, very primitive IDPs are present in nontargeted collections (see also Floss et al. 2006) and multiple fragments from cluster IDPs appear to be heterogeneous. Although high presolar grain abundances are indicative of primitive IDPs, presolar grain abundances alone cannot be used as a defining characteristic of primitiveness due to the highly heterogeneous distributions of presolar grains. The presence and abundance of other isotopic anomalies (such as D- and 15N-enrichments/depletions) alone are also unreliable indicators of primitiveness as IDPs with highly disordered organic matter and abundant presolar grains can lack such anomalies. This is, in part, because different phases respond differently to secondary processing. Raman spectroscopy may be one of the best indicators of primitiveness used alone but the most primitive meteorites also show extremely disordered organic matter (Busemann et al. 2007). Clearly, a combination of the above criteria is required for a thorough comparison of the relative degrees of primitiveness of IDPs.

The comparison of IDPs from targeted and nontargeted collections cannot yield a definitive classification scheme for the unambiguous identification of cometary material from specific sources. However, there are many observations that are unique to the targeted GSC IDPs (Busemann et al. 2009a), such as noble gases that imply extremely short free transfer times after ejection (Palma et al. 2005), the presence of an unknown, isotopically exotic 3He-rich and 22Ne-poor component (Palma et al. 2005), a new manganese-silicide phase (“Brownleeite”; Nakamura-Messenger et al. 2008, 2010), and large abundances of trapped Xe in metal-rich fragments (Busemann et al. 2010). Since targeted collectors are expected to contain still significant numbers of IDPs not originating from the desired source (Messenger 2002), many of which are likely from other cometary sources, a large amount of material of varying provenance will unavoidably be present. Non-GSC IDPs analyzed here appear to be more primitive than the GSC IDPs analyzed here in terms of presolar grain abundances and other isotope anomalies whilst the Raman spectral parameters appear to be more pristine in the GSC IDPs here and in those studied by Busemann et al. (2009b). The highly primitive nature of the very finely grained GSC IDP Drake is attested to by the fact that it is the only GSC IDP that contains any presolar grains (one silicate), it is enriched in D, and it has the most disordered OM of all the IDPs studied here. Therefore, this IDP may support the view that a number of particles collected in the dust stream of comet Grigg-Skjellerup are indeed fresh from this comet (Busemann et al. 2009a) as predicted by Messenger (2002). However, it is just as likely that it is from some other comet that happened to impact this collector. In light of the recent discovery of the hydrous nature of comets, a comet Grigg-Skjellerup origin for the hydrated GSC IDPs, Hawkins and Frobisher, cannot be ruled out. In fact, based on the observations of other particles from the same cluster (Nakamura-Messenger et al. 2008; Busemann et al. 2009b), Hawkins appears to represent a hydrated portion of a larger anhydrous cluster particle. Therefore, the presence of two hydrated GSC IDPs, one of which represents a hydrated portion of a larger anhydrous cluster IDP, out of three GSC IDPs analyzed may in fact reflect the hydrously altered nature of cometary material.


  • 1 NanoSIMS analysis of IDPs revealed the presence of abundant presolar silicate, oxide, SiC, and other grains with isotopically anomalous carbon, and other isotopic anomalies (such as D and 15N enrichments) associated with organic matter.
  • 2 Presolar grain abundances for the IDPs studied here are significantly higher than for even the most primitive meteorites, emphasizing the highly primitive nature of IDPs compared to other solar system materials.
  • 3 The diversity and abundances of presolar grains found in a relatively small sample set (10 IDPs) attests to the highly primitive nature of the majority of the IDPs studied here. These IDPs represent some of the most pristine solar system materials and are relatively unaltered compared to meteoritic material from the asteroid belt. The presence of N-isotopic anomalies and presolar SiC, other C-anomalous, silicate, and oxide grains, in one single IDP (Pizarro) testifies to the highly primitive nature of this particular IDP.
  • 4 The presolar silicates and oxides located here display a variety of O-isotopic compositions; similar to those from other IDPs and primitive meteorites. One grain appears to be the most 17O-rich presolar silicate ever found in an IDP (16O/17O = 156 ± 4; δ17O = 15700 ± 400‰), and one of the most 17O-rich in extraterrestrial materials reported to date. This grain is identified as a possible GEMS aggregate by TEM analysis (see also Davidson 2009).
  • 5 Hydrogen and nitrogen anomalies were present as so-called “hotspots” and “coldspots” (nitrogen only), testifying to the presence of unaltered OM from either the interstellar medium or the solar nebula. Some IDPs also have anomalous bulk compositions, including one of the few bulk 15N-depletions reported for an IDP (Hawkins; δ15N = −190 ± 27‰), resulting from more diffuse anomalies being present (as previously reported by Floss et al. 2006). No spatial correlation was found between H and N anomalies suggesting that they are not necessarily coupled.
  • 6 No relationship was seen between bulk N composition and primitiveness for the IDPs studied here, contradictory to criteria laid out by Floss et al. (2006) for quickly identifying the most primitive IDPs. Thus, using this criterion alone to identify “primitive” IDPs may bias the identification of the most primitive IDPs.
  • 7 Raman spectral parameters appear to offer a distinction between the GSC and non-GSC collections, particularly when combined with previous studies (Busemann et al. 2009), albeit still with considerable overlap between the two populations. IDPs from the same clusters appear to show similar degrees of primitiveness in terms of Raman spectral parameters, but highly variable presolar grain abundances (reflecting their heterogeneous natures).
  • 8 Data presented here suggest that IDPs incorporated heterogeneous mixtures of material even relative to one another (e.g., differences in bulk N composition between the IDPs studied here and by Floss et al. 2006).
  • 9 Of the three GSC IDPs studied here, one appears to be very primitive whilst another is a hydrated fragment of a larger anhydrous cluster particle. The presence of only one very primitive GSC IDP is statistically not surprising as between 1 and 50% of IDP flux at the time of the comet’s crossing with Earth were predicted to be from comet Grigg-Skjellerup (Messenger 2002). It is expected that these collectors contain abundant IDPs of nonspecific origin. However, many of these nonspecific IDPs should still be cometary in origin, and may be as primitive as or even more primitive than GSC IDPs, rendering the identification of a Grigg-Skjellerup comet origin very difficult.
  • 10 The majority of the non-GSC IDPs studied here are anhydrous, finely grained, and very primitive in nature. Although they are not as primitive as those reported as being potentially fresh from comet Grigg-Skjellerup (Busemann et al. 2009a) a cometary origin is still inferred for all anhydrous non-GSC IDPs.
  • 11 In terms of the data discussed above there appears to be no defining difference between the GSC and non-GSC IDPs analyzed here. However, this comparison is tentative due to the small sample set analyzed and inherent heterogeneities in IDPs at the micrometer scale. In future, the study of a large, unbiased number of IDPs from a variety of sources (i.e., associated and not associated with particular cometary dust streams) is needed to address this issue. However, given the nature of IDP studies this may be unreasonable and rather comparison of the recent number of successive studies (e.g., Floss et al. 2006, 2010; Busemann et al. 2009; this study) may be more viable.

Acknowledgments—  The authors would like to thank the Cosmic Dust AstroMaterials Curation Facility at JSC for providing samples, and Gordon Imlach at the OU for FE-SEM assistance. Larry Nittler is thanked for the use of L’IMAGE, helpful discussions, and inspiration. This paper benefitted from the editorial expertise and constructive comments of AE Christine Floss, and helpful suggestions and comments from three reviewers, Jan Leitner, George Flynn, and Jérôme Aléon. Funding from STFC is gratefully acknowledged for the PhD studentship that supported this research, the Aurora and advanced fellowship programme (HB), and for the UKCAN facilities employed throughout this work.

Editorial Handling—  Dr. Christine Floss