Pristine extraterrestrial material with unprecedented nitrogen isotopic variation

Abstract

Pristine meteoritic materials carry light element isotopic fractionations that constrain physiochemical conditions during solar system formation. Here we report the discovery of a unique xenolith in the metal-rich chondrite Isheyevo. Its fine-grained, highly pristine mineralogy has similarity with interplanetary dust particles (IDPs), but the volume of the xenolith is more than 30,000 times that of a typical IDP. Furthermore, an extreme continuum of N isotopic variation is present in this xenolith: from very light N isotopic composition (δ¹⁵N_AIR = −310 ± 20‰), similar to that inferred for the solar nebula, to the heaviest ratios measured in any solar system material (δ¹⁵N_AIR = 4,900 ± 300‰). At the same time, its hydrogen and carbon isotopic compositions exhibit very little variation. This object poses serious challenges for existing models for the origin of light element isotopic anomalies.

Physical and chemical conditions during the earliest stages of solar system evolution can be studied in chondritic meteorites and interplanetary dust particles (IDPs), believed to be among the most primordial objects left over from the formation of the solar system some 4.567 billion years ago (1). In this work we describe the unaltered mineralogy and light element (i.e., hydrogen, carbon, and nitrogen) isotopic composition of a primordial xenolith in the chondrite Isheyevo. Isheyevo is a Fe-Ni metal rich (60 vol% on average) chondrite with 2 dominant lithologies that resemble CB and CH chondritic material (2). Apart from Fe-Ni metal grains, this rock is composed of chondrules, rare Ca-Al rich inclusions (CAIs) and hydrated lithic clasts but does not otherwise contain fine-grained matrix, which separates high-temperature components in most other types of chondritic materials (2–4).

In a survey of 2 polished sections of Isheyevo (the CH lithology), representing a total area of about 400 mm², we identified more than 100 lithic clasts. Their mineralogy has been studied by energy dispersive X-ray (EDX) spectroscopy, analytical electron microprobe (AEM), scanning electron microscopy (SEM), and transmission electron microscopy (TEM). In general, these lithic clasts show a high degree of aqueous alteration, consistent with previous reports (3, 4). This indicates that they are not genetically related to the high-temperature components in Isheyevo, which do not show effects of in situ aqueous alteration (3, 4).

Lithic clasts found in our 2 sections range in size between 50 and 750 μm. Their mineralogical properties are broadly similar to matrix material of CI, CM, or CR carbonaceous chondrites, that is, dominated by a fine-grained matrix composed of sub-μm–size phases, mainly phyllosilicates with variable amount of Fe-Ni sulfides and magnetite. Matrix AEM point analyses show high contents of S (between 2 and 15 wt%), due to the presence of either tochilinite and/or Fe-Ni sulfides. Anhydrous silicates such as olivine and pyroxene, present either as isolated crystals in the matrix or as microchondrules (≈30–60 μm in diameter), are less abundant. Olivine is mainly Mg-rich, with compositions ranging from Fo₉₈ and Fo₈₄, with a mean value of Fo₈₉ [Fo is defined as the mol% ratio Mg/(Mg+Fe)]. Pyroxene ranges in composition between En_63–98 Wo_0–4 Fs_1–36 [En = Mg/(Mg+Fe+Ca) mol%; Wo = Ca/(Mg+Fe+Ca) mol%; Fs = Fe/(Mg+Fe+Ca)]. Magnetite is very abundant in some lithic clasts and absent in others. Carbonates are rare but have been identified in some cases. Sulfide grains are mostly pyrrhotite, with some rare grains of pentlandite. Metal is present as small, μm-sized grains associated with sulfides.

Among the approximate 100 lithic clasts studied here, 1 stood out by its distinct textural and mineralogical properties and for its extreme variation in N isotopic composition. We describe here the discovery of this primordial xenolith (hereafter named PX-18) and the implications concerning models for light element fractionation.

PX-18: Mineralogy

PX-18 is a dark xenolith (≈380 × 470 μm²), dominated by a very fine-grained matrix, mainly composed of anhydrous Mg-rich silicates with tiny Fe-Ni sulfides grains and magnetite (Fig. 1A). High-resolution SEM images and EDX spectral analyses revealed a matrix composed of sub-μm-sized crystals, including Mg-rich olivine and pyroxene crystals (Fo_84–89; En_89–95, Wo_0–4) with sizes of about 1–2 μm, Fe-Ni metal grains, magnetite, and very rare carbonates grains.

Fig. 1.

PX-18 in the Isheyevo chondrite. (A) High-resolution SEM image. The minimum and maximum dimensions of PX-18 are 380 μm and 470 μm, respectively. This xenolith is clearly dominated by matrix, but some large olivine and pyroxene grains are also visible. White inclusions are metal grains. The black arrow indicates the point from which the TEM thin section showed in B has been extracted. (B) TEM bright-field micrograph of a thin section extracted by focused ion beam (FIB) technique from PX-18 in Isheyevo. Individual analyses made by TEM are reported in the figure. A large olivine grain (Fo₁₀₀) is visible at the bottom of the section. “En” labels indicate pyroxene grains with En₉₈₋₁₀₀. The other pyroxene grains analyzed are: px1 (En₈₇ - Wo₁); px2 (En₉₁ - Wo₁); px3 (En₆₃ - Wo₁); px4 (En₉₅ - Wo₁); and px5 (En₉₃ - Wo₁). Fe-Ni metal grains (“met”) are visible in the left part of the section. Two grains of pyrrhotite are also present (“pyr”). Labels “mx” indicate a very fine-grained matrix, composed of tiny crystals and minor amorphous material, with average composition frequently close to olivine Fo₆₀.

TEM examination confirmed that the PX-18 matrix is primarily composed of crystalline, anhydrous silicate grains, mainly Mg-rich pyroxene, with a few, μm-sized forsterite (Fo_97–99) grains (Fig. 1B). Low-Ca pyroxene grains show a wide compositional range (En_63–100, Wo_0–5) but enstatite clearly dominates (36% of analyzed grains has En_99–100). The most fine-grained phase (labeled “mx” in Fig. 1B) contains abundant Fa-rich olivine (grain size 20–100 nm) with compositions ranging from Fo₂₄ to Fo₇₀, and an average around Fo₄₀. Small Fe-Ni metal and Fe-oxide grains are present as minor components between silicates. Importantly, high-resolution TEM observations did not did not yield any occurrence of layered phyllosilicates, which seem to be absent from PX-18. These observations indicate that PX-18 is mineralogically similar to other primitive material, such as chondritic porous IDPs (5) and comet 81P/Wild2 samples returned by the Stardust mission (6). Clearly, PX-18 has avoided the extensive aqueous alteration that affected the other xenoliths in Isheyevo (2) and therefore, represents a more pristine sample, which might have better preserved isotopic anomalies of light elements.

PX-18: Isotopic Composition

We performed isotopic composition analyses of PX-18 and 4 other xenoliths, covering a total area of 30,400 μm². The isotopic ratios D/H, ¹³C/¹²C, and ¹⁵N/¹⁴N were imaged by multicollection secondary ion mass spectrometry, using a CAMECA NanoSIMS N50. We report our results in the δ-notation, which expresses the deviation in parts per thousand from terrestrial standards (i.e., the SMOW value for the D/H ratio, the atmospheric ¹⁵N/¹⁴N ratio for N, and the PDB value for ¹³C/¹²C).

NanoSIMS imaging of the H isotopic composition in the 5 xenoliths yielded homogeneous distributions with average δD_SMOW values ranging between −70 ± 20‰ and 420 ± 50‰. In other words, the bulk D/H ratios obtained are generally higher than the SMOW value, but no extreme D/H ratios were observed, neither in bulk nor as individual hotspots, compared with previous studies of IDPs (7) and insoluble organic matter (IOM) extracted from primitive chondritic meteorites (8). Carbon isotopic compositions were also found to be relatively homogeneous within individual xenoliths with values ranging from δ¹³C_PDB = −35 ± 3‰ to −17 ± 2.5‰, comparable to bulk C isotopic composition of chondritic meteorites (9) and IOM (10).

The N isotopic compositions yielded a dramatically different data set. For all 5 xenoliths, systematic bulk enrichments in ¹⁵N were observed, with δ¹⁵N_AIR ranging from 110 ± 10‰ to 318 ± 16‰. Even larger N isotopic anomalies were found in the mineralogically primitive PX-18. The main characteristics of the N isotopic distribution in PX-18 are the following: 1) large areas with a diffuse, but substantial enrichment in ¹⁵N over the surrounding material (Fig. 2A); 2) extremely ¹⁵N-enriched, localized anomalies (hereafter called hotspots, Fig. 2B); and 3) areas with negative δ¹⁵N_AIR values, comparable to those of the solar nebula (11–13).

Fig. 2.

Results of NanoSIMS measurements on PX-18: the δ¹⁵N_AIR distribution of two 40 × 40 μm² regions (A and B) and hotspots with internal structures (C–E). Color scale has been changed for images C to E to better show internal structures of hotspots. In all of the images, the black areas are regions in which precise isotopic ratios could not be determined. These pixels have therefore been removed from each image. (A) Region with average δ¹⁵N_AIR = 640 ± 11‰. About 90% of this image is characterized δ¹⁵N_AIR greater than or equal to 250‰ (Table 1). Several hotspots are also visible (indicated by white arrows). (B) Area with average δ¹⁵N_AIR = 700 ± 17‰. Two hotspots, approaching δ¹⁵N_AIR = 4000‰, are indicated by arrows. (C) The hotspot is 0.598 μm² (≈773 × 773 nm²) in size and has δ¹⁵N_AIR = 3,200 ± 150‰. The yellow internal region has a surface of 0.251 μm² (≈500 × 500 nm²) and δ¹⁵N_AIR = 4,110 ± 200‰. ¹⁵N-richer, white areas are visible. They have δ¹⁵N_AIR = 4,900 ± 300‰, and a total surface of 0.0967 μm² (≈310 × 310 nm²). This is the only case in which is possible to define 2 internal structures in the same hotspot. (D) For the larger isotopic anomaly (1.29 μm², δ¹⁵N_AIR = 2,600 ± 170‰) an internal hotspot corresponding to the yellow region can be defined, with δ¹⁵N_AIR = 3,700 ± 300‰ and a surface of 0.467 μm² (≈683 × 683 nm²). For the smaller isotopic anomaly (1.1 μm², δ¹⁵N_AIR = 1700 ± 110‰), the internal hotspot has an area of 0.124 μm² (≈352 × 352 nm²), with δ¹⁵N_AIR = 3,000 ± 300‰. (E) Again, the yellow region represents the internal hotspot, corresponding to δ¹⁵N_AIR = 3,100 ± 470‰ with a surface of 0.079 μm² (≈280 × 280 nm²) [for the entire hotspot the size is 0.896 μm² (≈946 × 946 nm²) and δ¹⁵N_AIR = 1,800 ± 160‰].

A total of 13 NanoSIMS images were obtained from PX-18, covering an area of approximately 12,800 μm². These images have average δ¹⁵N_AIR values that range between 30 ± 20‰ and 700 ± 20‰, indicating a highly heterogeneous distribution of the ¹⁵N/¹⁴N ratio in PX-18 (Table 1). In 3 images obtained on PX-18 (each 40 × 40 μm²), more than 90% of the surface is characterized by δ¹⁵N_AIR greater than or equal to 250‰ and very elevated average δ¹⁵N_AIR values (630 ± 20‰, 640 ± 11‰ and 700 ± 17‰). These high values are not due to the presence of only a few small anomalously ¹⁵N-rich regions but rather to the occurrence of a diffuse ¹⁵N-rich component, present in large fractions of the images (Fig. 2 A and Table 1). This is demonstrated by the frequency distribution of δ¹⁵N_AIR values obtained from a grid of 600 × 600 nm² squares laid over each δ¹⁵N_AIR image. The resulting frequency distributions (Fig. 3) clearly show continuous and broad ranges of δ¹⁵N_AIR values, centered on the δ¹⁵N_AIR average value obtained for each image.

Table 1.

Characteristics of isotopic measurements for eight 40- × 40-μm² images from PX-18

Image	Area, μm2	Average δ15NAIR, ‰	δ15NAIR > 250‰, % area	δ15NAIR < 0‰, % area	No. of hotspots
g_31	394.6	140 ± 13	34.26	27.42	3
g_32	541.2	630 ± 20	92.63	1.05	10
g_33	507.6	700 ± 17	95.76	0.79	2
g_41	1218.9	220 ± 9	35.89	29.95	7
g_42	1082.1	640 ± 11	89.67	1.45	16
g_48	888.6	110 ± 20	23.48	43.18	7
g_49	601.4	30 ± 22	28.78	52.63	0
g_53	995.8	80 ± 17	28.26	46.67	1

The column labeled “Area” reports the fraction of each ratio image that has a number of ¹⁴N counts at least equal to 5% of the ¹⁴N-image maximum. Average δ¹⁵N_AIR values are evaluated only on these image portions. These same regions also define for each image the basic area to which percentages of δ¹⁵N_AIR > 250 ‰ and δ¹⁵N_AIR < 0 ‰ refer.

Fig. 3.

Frequency distribution for δ¹⁵N_AIR values of three 40 × 40 μm² images from PX-18. These distributions are obtained from a grid of 600 × 600 nm² regions of interest (ROIs) laid over each NanoSIMS image in areas for which ¹²C¹⁴N signal was at least 5% of the image maximum. See Table 1 for average values of each image. Vertical scale is the number of square ROIs in each δ¹⁵N bin. The right-hand tails of these distributions are due to hotspots, but the correspondence between ROIs and hotspots is only partial, because the randomly distributed, 600 × 600 nm², ROIs do not exactly match hotspots. (A) Image g_32, the distribution is obtained from a total surface of 410.45 μm². (B) Image g_33, the distribution is obtained from a total surface of 325.5 μm². The distribution also contains a hotspot at δ¹⁵N_AIR = 3,900‰ (visible in the Inset). (C) Image g_42, the distribution is obtained from a total surface of 639.45 μm².

In addition to the diffuse distribution of ¹⁵N-enriched material, forty-six ¹⁵N hotspots with extremely high δ¹⁵N_AIR were observed in PX-18 (Fig. 2B). These hotspots, with areas of approximately 1 μm², are distinct from the aforementioned, broad ¹⁵N-enriched zones, because their δ¹⁵N_AIR values minus a 3σ error bar (where σ is the error on the hotspot δ¹⁵N_AIR value) are higher than the average δ¹⁵N_AIR value for the image plus 3σ_av (where σ_av is the error on the average δ¹⁵N_AIR value of the entire image). These hotspots are highly heterogeneously distributed. For example, 1 NanoSIMS image (40 × 40 μm²) contains no hotspots, while other areas in PX-18 of similar size have between 2 and 10 hotspots, and 1 image contains as much as 16 hotspots (Table 1). All of the ¹⁵N hotspots have sizes larger than 300 × 300 nm², up to approximately 2 × 2 μm². The highest δ¹⁵N_AIR for a single hotspot is 3,200 ± 150‰. However, several hotspots exhibit distinct internal structures, that is, small internal regions with even higher values of δ¹⁵N_AIR. Four such cases have been identified (Fig. 2 C–E), in which the averages δ¹⁵N_AIR values for each hotspot are 1,700 ± 110‰, 1,800 ± 160‰, 2,600 ± 170‰, and 3,200 ± 150‰, and in which there are subregions with higher δ¹⁵N_AIR values of 3,000 ± 300‰, 3,100 ± 500‰, 3,700 ± 300‰, and 4,900 ± 300‰, respectively. These hotspot subregions are the highest δ¹⁵N_AIR values ever measured in solar system material: 3,700 ± 300‰ and 4,900 ± 300‰ (Fig. 4). We stress here that these are results obtained by in situ measurements, and hence they represent an isotopic composition that is a mix of the ¹⁵N-rich phase with other, ¹⁵N-poor phases. This implies that our highest δ¹⁵N_AIR values are a lower limit for the ¹⁵N enrichment in the carrier phase.

Fig. 4.

δ¹⁵N_AIR values measured in PX-18, compared with other data for solar system materials. Ranges reported in the figure comprise results from bulk measurements as well as from hotspots. The mean δ¹⁵N_AIR for the whole PX-18 is 318 ± 16‰. Bulk δ¹⁵N_AIR values for the other materials reported in the figure range between −52.5‰ and 2,500‰ for CB/CH chondrites (31–33), between −108 ± 9‰ and 500 ± 20‰ for IDPs (7), between −54 ‰ and 452‰ for Wild2 samples [more negative values are due to the presence of ¹⁵N-poor, ¹³C-rich hotspots (34)], and between −66 ± 3‰ and 415 ± 2‰ for IOM extracted from carbonaceous chondrites (8, 10). The in situ measurements, and those on IOM, have been obtained by SIMS techniques. Stepped combustion has been performed for CB/CH chondrites. For HCN molecules in comets, astronomical spectroscopy measurements (35) yielded values of δ¹⁵N_AIR = 330 ± 450‰ (for comet Hale-Bopp) and 960 ± 370‰ (for comet 17P/Holmes). For CN molecules in comets, δ¹⁵N_AIR values range between 650 ± 400‰ and 1,100 ± 450‰ (35, 36). The values of the N isotopic composition for the solar nebula are obtained from measures on: 1) ammonia in the Jupiter atmosphere analyzed by the Galileo spacecraft [δ¹⁵N_AIR = −370 ± 80‰, (13)]; 2) solar wind N implanted in lunar regolith grains [δ¹⁵N_AIR ≤ −240‰, (12)]; 3) an osbornite (TiN) inclusion in a CAI of Isheyevo [δ¹⁵N_AIR = −359 ± 5‰, (11)].

Regions with negative δ¹⁵N_AIR values have been also identified in PX-18. In some cases, they represent up to 52% of the analyzed surface (Table 1). The global mean δ¹⁵N_AIR value evaluated from all of the image fractions with negative δ¹⁵N_AIR is −150 ± 30‰, and the minimum value measured for a single image is δ¹⁵N_AIR = −310 ± 20‰ (Fig. 4). This latter value is consistent with values inferred for the solar nebula (11–13) from previous measurements (see Fig. 4 for details).

The large range of δ¹⁵N_AIR values observed in PX-18 from Isheyevo (Fig. 4) greatly expands the range of N isotopic composition for a single extraterrestrial object or type of material (e.g., IOM). On one hand, the heaviest N isotopic compositions observed (δ¹⁵N_AIR = 3,700 ± 300‰ and 4,900 ± 300‰) are enriched in ¹⁵N by factors of 4.7 and 5.7, respectively, compared with terrestrial atmospheric N and even more (factors of 7.3 and 9.1, respectively) when compared with the best estimate for the N isotopic composition of the solar nebula (11). On the other hand, negative δ¹⁵N_AIR values (δ¹⁵N_AIR = −310 ± 20‰) similar to those of the solar nebula and corresponding to the lowest observed values, have also been measured. This, combined with PX-18 unique mineralogy, is consistent with the notion that only the most pristine, unaltered material can preserve primordial isotopic compositions in light elements.

The Carrier Phase of the N Anomalies

To verify whether ¹⁵N hotspots in PX-18 are associated with specific carriers, it is possible to exploit the high spatial resolution of the NanoSIMS images (100–200 nm). By comparing NanoSIMS images with high resolution SEM images (see Fig. S1), the site of the ¹⁵N hotspots in PX-18 were located in the matrix and studied in more detail. Several important observations were made: (i) hotspots have linear dimensions substantially larger than the size of typical matrix grains (see Fig. 5 for an example), (ii) typically, hotspot regions contain abundant, sub-μm-sized grains of Mg-rich pyroxene and magnetite, and (iii) EDX spectra of hotspot regions show a general enrichment in C with respect to spectra obtained in other regions of PX-18 (Fig. 5).

Fig. 5.

High-resolution EDX analyses of PX-18 matrix. (A) High-resolution SEM image of a PX-18 matrix area corresponding to 1 hotspot, indicated by the ellipse. Crosses labeled with white numbers are the points analyzed by EDX spectroscopy. (B and C) EDX spectra of points number 1 and 5, respectively. The first peak on the left is that of carbon, clearly evident in these 2 spectra. (D) EDX spectrum of point number 3. In this case, the C peak is smaller, corresponding to the background signal due to the section carbon coating.

Presolar SiC grains or nanoglobules (14) can be excluded as the carrier phase of the N isotopic anomalies because a) they generally have smaller size than the diffuse ¹⁵N enrichments observed here, b) if ¹⁵N hotspots were due to presolar grains or nanoglobules, they would be identified in high resolution SEM images at the hotspots locations (or in TEM observations), c) organic nanoglobules have D enrichments (14), not measured in PX-18, and d) most SiC grains carry with them large anomalies in C, which are not observed (15). Together, these observations lead to the conclusion that ¹⁵N hotspots in PX-18 are due to the presence of organic matter (OM).

The spatial extension of ¹⁵N-rich regions in PX-18 suggests the presence of a diffuse component as carrier of these anomalies. Indeed, macromolecular OM has been identified in PX-18 by Raman spectroscopy. PX-18 was analyzed together with matrix fragments extracted from pristine chondrites that were used as standards: Murchison (CM2), Cold Bokkeveld (CM2), Semarkona (LL/3.00), and Bishunpur (LL/3.15). The spectra exhibit the first-order carbon bands D and G, superimposed onto a fluorescence background indicating that the macromolecular OM is widespread and abundant. Inspection of the combined set of spectra (Fig. S2) shows that in PX-18 these 2 bands are wider and more overlapping than in the type 3 ordinary chondrites Semarkona and Bishunpur. This suggests a lower degree of thermal metamorphism in PX-18 macromolecular OM.

The Lorentz-Breit-Wigner-Fano (LBWF) fit analysis confirmed the visual analysis of the spectra. The values of the spectral parameter FWHM-D (D band full width at half maximum), which is a sensitive tracer of thermal metamorphism (16), rule out the presence of significant thermal metamorphism. Raman spectra from PX-18 have mean FWHM-D of 268 ± 8 cm⁻¹ (Fig. S3), a value between those of the metamorphosed chondrite Bishunpur (170 ± 20 cm⁻¹) and Semarkona (250 ± 3 cm⁻¹) and those of the little metamorphosed CM2 Cold Bokkeveld (282 ± 12 cm⁻¹). To characterize highly disordered carbonaceous matter, the most suitable spectral parameters are FWHM-G and ω_G (G band peak position). A rough comparison can be made with data from IDPs and chondrites (17), keeping in mind that different experimental conditions have been used. The G-diagram (Fig. 6) indicates that the macromolecular OM in PX-18 has a degree of thermal metamorphism lower than that of the Bishunpur and Semarkona chondrites and comparable to that of organic matter present in IDPs.

Fig. 6.

G-diagram for results of Raman analyses on PX-18: the spectral parameter FWHM-G plotted against the peak position ω_G. This diagram reflects the degree of thermal metamorphism and structural order of the organic matter. The solid and dashed lines show the range of values for chondritic meteorites and IDPs respectively (17). CM measurements were obtained from matrix of Murchison (CM2) and Cold Bokkeveld (CM2).

Nitrogen Fractionation

Excluding a stellar nucleosynthesis origin (i.e., related to presolar grains) for the observed N isotopic anomalies, processes such as self-shielding and low-temperature ion-molecules reactions can be considered. Self-shielding (18) is still being explored as an explanation for O fractionation in early solar nebula (19–21) and it has been proposed to be effective also for N in the protosolar accretion disk (20). However, quantitative models for N fractionation due to self-shielding have not been developed and the extent of N isotopic fractionation by this process is unknown.

Models for isotopic fractionation by low-temperature ion-molecule reactions can produce values of δ¹⁵N_AIR as high as those observed in PX-18. In the most recent model (22), which invokes chemical reactions with N-containing molecules under dark molecular cloud conditions, values for δ¹⁵N_AIR of >9,000‰ for external layers of NH₃ ice accreted on dust grains are obtained. Transfer of fractionated N from NH₃ ice to organic matter is achieved by UV-induced transformations into polycyclic aromatic hydrocarbons (PAHs). PAHs are abundant in the gas phase where N fractionation take place and they can condense on dust grains and form ice (23). Experiments on various coronene-ice mixtures at low temperature (15 K) and low pressure have shown that -NH₂ functional groups present in the ice can be added to PAHs by the action of UV photolysis (23).

However, a fundamental problem is that low temperature ion-molecule reactions are also predicted to produce strong deuterium enrichments in organic matter (24, 25), which are not found in PX-18 or any other xenolith in Isheyevo. Other previous studies (e.g., 2, 9, 10) have also found that H and N anomalies are decoupled in extraterrestrial matter. These results call for a new theoretical and experimental approach, which must be able to provide an explanation for the decoupling of these light element isotopic variations.

The primitive mineralogy of PX-18, in combination with the extreme range of N isotopic variation and the un-metamorphosed organic matter it preserves, indicate that PX-18 represents a class of extremely pristine extraterrestrial material, which might sample little processed outer solar system bodies. A Kuiper Belt parent body is a likely possibility. The presence of outer solar system material in an asteroidal body is a natural consequence of the idea that xenoliths represent samples of primordial matter scattered throughout the solar system during the Late Heavy Bombardment (26). In the context of global scattering and mixing of bodies (27), objects composed of more fragile and/or unconsolidated material, such as primordial chondritic planetesimals, comets, and Kuiper Belt objects, are greatly affected by collisions (28), and can produce fragments that are scattered throughout the solar system and become embedded in meteorite parent bodies.

Materials and Methods

To search for primordial xenoliths, preliminary SEM and EDX spectral analyses of 2 Isheyevo polished sections have been performed at the Laboratoire de Minéralogie et Cosmochimie du Muséum of the Muséum National d'Histoire Naturelle (Paris, France) by mean of a JEOL JSM 840-A SEM equipped with an EDAX Genesis X-rays detector, using a 3-nA primary beam accelerated by a 15-kV potential difference. High-resolution SEM images have been acquired at the University of Paris VI using a Zeiss Supra55 VP field emission SEM, with a 4-nA, 10-kV accelerated primary beam. By the same instrument, EDX spectral analyses of sub-μm-sized grains in the matrix of PX-18 have been performed.

Quantitative analyses for the mineralogical composition of xenoliths have been obtained with a CAMECA SX-100 electron microprobe at the University of Paris VI. A 10-nA focalized beam, accelerated by a 15-kV potential difference, is used for point analyses of xenoliths silicates, oxides, and metals.

For TEM analyses 2 thin sections have been extracted from PX-18 by focused ion beam technique at the Institut d'Electronique, de Microélectronique et de Nanotechnologie (Lille, France). Individual matrix grains have been analyzed by a Philips CM 30 at the Laboratoire de Structure et Propriétés de l'Etat Solide of the Université des Sciences et Technologies (Lille, France). The accelerating voltage was 300 kV. EDX microanalyses have been realized by a Noran detector.

Raman measurements have been performed at the Laboratoire de Sciences de la Terre (ENS) with a Labram spectrometer (Horiba-Jobin-Yvon), equipped with a Spectra Physics Argon ion laser using 514.5 nm excitation. The laser beam was focused by a microscope equipped with a long distance working objective (X50), leading to a 2–3 μm spot. The reproducibility of Raman measurements on poorly ordered macromolecular organic matter is controlled by atmospheric conditions, irradiation time, and power on the sample, due to heating and photo-oxidation effects during irradiation (29). Hence spectra have been acquired under strictly constant experimental conditions: power at the sample surface was 400 ± 5 μW, exposure time of each measurement was 90 s and under an inert argon atmosphere. A 600 gr/mm grating has been used and provided with spectra in the spectral region 500–2,200 cm⁻¹, recovering the first- and second-order carbon bands. Raman spectra have been quantitatively analyzed by fitting the first-order carbon bands using a Lorentzian profile for the D peak, and a Breit-Wigner-Fano one for the G peak [LBWF fit: see (30) for a discussion regarding the choice of these profiles]. The spectral parameters derived from the LBWF fit are the width of the bands FWHM-G and FWHM-D and their peak position ω_G and ω_D.

Secondary ion mass spectrometry analyses were performed with a NanoSIMS CAMECA N50 at the Laboratoire de Minéralogie et Cosmochimie du Muséum of the Muséum National d'Histoire Naturelle of Paris. We have used a Cs⁺ primary beam in two sessions. The first session for detection of H⁻ and D⁻ (with a current on the sample of ≈40 pA) and the second one for ¹²C⁻, ¹³C⁻, ¹²C¹⁴N⁻, and ¹²C¹⁵N⁻ (current on the sample ≈4 pA). In each case an electron beam has been used to compensate for charging effects. For H isotope measurements, the experimental set up was such that the mass resolution was ΔM/M greater than or equal to 2,000. For C and N isotopes, a larger mass resolution is needed, so the machine was set up to yield ΔM/M between 7,500 and 8,000. This is necessary to prevent possible interferences between different isotope masses, for example, ¹¹B¹⁶O⁻ on ¹²C¹⁵N⁻. To determine the instrumental mass fractionation (IMF) we have used as standard reference a type III kerogen, with known isotopic compositions of H, C, and N in the same reference sample. The IMF was on the order of −315 ± 5‰ for H, −50 ± 1‰ for C, and −6 ± 4‰ for N. We selected for isotopic analysis a few 40 × 40 μm² matrix-dominated regions from 5 xenoliths of Isheyevo. These regions were exposed to high Cs⁺ current (≈1 nA) for 15–20 min before analyses, to reach sputtering equilibrium. Measurements were made in scanning imaging mode. For each selected region, between 7 and 35 plans composed of 256 × 256 pixels were collected. Measurement time was 65.536 seconds per plan in the case of H isotopes and 327.68 seconds per plan for C and N isotopes. Where hotspots were found, close-up analyses on ≈10 × 10 μm² areas have been performed. Details for data reduction can be found in the on-line SI Text. Here we stress our criteria to define hotspots in ¹⁵N/¹⁴N ratio images. A selected region of interest (ROI) in 1 image is defined as a hotspot if: 1) (δ¹⁵N_AIR)_hotspot − 3σ_hotspot > (δ¹⁵N_AIR)_av + 3σ_av, where σ_hotspot is the error on the (δ¹⁵N_AIR)_hotspot value of the selected ROI and (δ¹⁵N_AIR)_av and σ_av are the average δ¹⁵N_AIR value and error for the entire image; 2) the hotspot is larger than the image spatial resolution (all potential hotspots with size smaller than 250 × 250 nm² have been discarded); and 3) each individual plan that composes the image has a ¹⁵N/¹⁴N ratio that differs no more than 3σ_mean from the mean ¹⁵N/¹⁴N ratio (where σ_mean is the standard deviation calculated from the all set of isotopic ratios relative to the same ROI).

The errors reported in the text express a 2σ confidence level.