Abstract– Asteroid 2008 TC3 impacted Earth in northern Sudan on October 7, 2008. The meteorite named Almahata Sitta was classified as a polymict ureilite. In this study, 40 small pieces from different fragments collected in the Almahata Sitta strewn field were investigated and a large number of different lithologies were found. Some of these fragments are ureilitic in origin, whereas others are clearly chondritic. As all are relatively fresh (W0–W0/1) and as short-lived cosmogenic radioisotopes were detected within two of the chondritic fragments, there is strong evidence that most, if not all belong to the Almahata Sitta meteorite fall. The fragments can roughly be subdivided into achondritic (ureilitic; 23 samples) and chondritic lithologies (17 samples). Among the ureilitic rocks are at least 10 different lithologies. A similar number of different chondritic lithologies also exist. Most chondritic fragments belong to at least seven different E-chondrite rock types (EH3, EL3/4, EL6, EL breccias, several different types of EL and EH impact melt rocks and impact melt breccias; some of the latter are shock-darkened). In addition, two H-group ordinary chondrite lithologies were identified, and one sample of a chondrite type that is so far unique. The latter has some affinities to R chondrites. Oxygen isotope compositions of 14 fragments provide further fundamental information on the lithological heterogeneity of the Almahata Sitta meteorite. Based on the findings presented in this study, the reflectance spectrum of asteroid 2008 TC3 has to be evaluated in a new light.
Asteroid 2008 TC3 was the first asteroid detected in space before impacting Earth and fell in the Nubian Desert of northern Sudan on October 7, 2008. Hundreds of mostly small fragments were recovered. The meteorite called Almahata Sitta was classified as a polymict ureilite (Jenniskens et al. 2009) based on the study of only one meteorite fragment from the strewn field. The analysis by these authors shows that this meteorite is “an achondrite, a polymict ureilite, anomalous in its class: ultra-fine-grained and porous, with large carbonaceous grains.” We have studied 40 small pieces from different fragments collected in the Almahata Sitta strewn field and found a large number of different lithologies. Some of these fragments are ureilitic in origin, whereas others are clearly chondritic. As all are relatively fresh (W0–W0/1; see the Samples and Analytical Techniques section), there are strong indications that most, if not all belong to the Almahata Sitta meteorite fall.
In this study, we will present mineralogical and oxygen isotope data for several Almahata Sitta meteorite fragments and the results of a study of short-lived cosmogenic radioisotopes of two chondritic fragments. Some preliminary data have been published by Bischoff et al. (2010) and Horstmann and Bischoff (2010a).
Samples and Analytical Techniques
All 40 studied Almahata Sitta fragments were found as individual meteorite fragments of 2.6–50 g in the Almahata Sitta strewn field in two search campaigns about 9 and 12 months after the meteorite fall. Most samples have a complete black to dark-brownish fusion crust. The partly brownish taint of several samples (e.g., MS-CH, MS-16, MS-152, and MS-165) is certainly the effect of desert weathering that occurred between the fall and the recovery. As all samples contain significant abundances of metal, the weathering grade in their interior can easily be determined. Based on Wlotzka (1993), no sample has a weathering degree higher than W0/1 as indicated by a slight brownish taint in transmitted light and/or by only very thin rinds of weathering products surrounding the metal grains (Tables 1 and 2). This fresh appearance is well-documented for 12 of the 40 samples in Figs. 1–5, where metals are shown in reflected light or backscattered electron (BSE) images.
Table 1. Ureilitic lithologies in the Almahata Sitta meteorite.
Fragment mass (g)
Typical grain sizec
Ol cores (Fa)
Px cores (Fs)
Px cores (Wo)
Note: W = weathering; Ol = olivine; Px = low-Ca pyroxene; recry. Ol = recrystallized olivine.
bWeathering grade W0/1 (Wlotzka 1993) assigned based on the slight brownish taint in restricted areas of the thin section as observed in transmitted light.
Oldhamite, troilite, keilite, Zn-alabandite
14.4 ± 7.5
Mainly: ∼38; Co: ∼2
Fa range: 19–38; matrix: ∼40 vol%
Oldhamite, troilite, keilite
Co in metal: ∼0.8; shock-melted areas
0.6 ± 0.4
Oldhamite, troilite, niningerite, keilite
2.9 ± 3.7
Oldhamite, troilite, daubreelite, niningerite
0.2 (1 grain)
0.5 ± 0.3
Oldhamite, (Zn-)alabandite, troilite, keilite
Contains shock-melted areas
Troilite, oldhamite, alabandite
Troilite, oldhamite, alabandite
Oldhamite, troilite, keilite
Co in metal: ∼0.7; shock-darkened
0.38 ± 0.2
Oldhamite, troilite, niningerite, keilite
0.4 ± 0.3
Oldhamite, troilite, keilite, one grain of niningerite
Breccia; niningerite-bearing (EH) fragment?
1.1 ± 0.55
Troilite, keilite, niningerite, oldhamite
0.4 ± 0.5
Schreibersite, contains shock-melted areas
Troilite, keilite, oldhamite, alabandite
Troilite, oldhamite, alabandite, keilite
A small piece of each of these fragments was selected for thin section preparation. Two sliced or partly sliced fragments of 5.1 g (MS-CH) and 8.65 g (MS-D) were used to measure the short-lived nuclides at the Laboratori Nazionali del Gran Sasso (LNGS, Italy; Tables 3 and 4). Small grains from several different fragments (<5 mg each) were prepared for oxygen isotope measurements at the Universität Göttingen (Germany).
Table 3. Data summary for the detected cosmogenic radionuclides in two samples of the Almahata Sitta meteorite. The reported uncertainties in the last digits (in parentheses) are expanded uncertainties with k = 1, the upper limits are given with 90% CL. For data on MS-CH, see also Horstmann et al. (2010).
Activity concentrations in [dpm kg−1]
MS-CH (5.1 g)
MS-D (8.65 g)
Table 4. Results for the naturally occurring nuclides Th, U, and Knat in two samples of the Almahata Sitta meteorite. The reported uncertainties in the last digits (in parentheses) are expanded uncertainties with k = 1, the upper limits are given with 90% CL.
Concentrations in [ng g−1]
MS-CH (5.1 g)
MS-D (8.65 g)
635 (100) × 103
632 (66) × 103
The mineralogy and texture of the fragments were studied by light and electron optical microscopy. A JEOL 6610-LV electron microscope was used to resolve the fine-grained textures and to analyze the mineral constituents using the EDS attached (INCA; Oxford Instruments). As natural standards we used olivine (Mg, Fe, Si), jadeite (Na), plagioclase (Al), sanidine (K), diopside (Ca), rutile (Ti), chromite (Cr), rhodonite (Mn), and pentlandite (Ni).
The oxygen isotope compositions of 14 fragments were measured by laser fluorination gas mass spectrometry. Sample material (typically ∼1–2 mg) is reacted with purified F2 gas with aid of a 50 W infrared laser. Excess F2 is reacted to Cl2 in a NaCl trap and Cl2 is trapped in a cold trap (−196 °C). Sample O2 is analyzed in continuous flow mode with a ThermoElectron MAT 253 gas mass spectrometer. Accuracy and precision in δ18O and Δ17O are typically ±0.2‰ and ±0.06‰, respectively. The terrestrial fractionation line (TFL) is defined by 290 analyses of rocks and minerals to β = 0.5250 ± 0.0007 (1σ).
The short-lived cosmogenic radioisotopes of two chondritic fragments from the Almahata Sitta strewn field were measured by means of gamma ray spectroscopy. The measurements were performed using high-purity germanium (HPGe) detectors, in ultra-low-background configuration (25 cm of lead and an inner liner of 5 cm copper, inside an underground laboratory with 1400 m rock overburden). The counting efficiency was determined with a thoroughly tested Monte Carlo code. The samples were measured from October 13 to 18, 2009 in the case of the sample MS-CH, and from November 9 to December 6, 2009, in the case of the sample MS-D.
In this study, we will present mineralogical and oxygen isotope data, as well as the results of a cosmogenic radioisotope study on various meteorite fragments from the Almahata Sitta strewn field. The fragments can roughly be subdivided into achondritic (ureilitic; 23 samples) and chondritic lithologies (17 samples). Main mineralogical characteristics are given in Tables 1 and 2. Details on some fragments have been previously presented by Bischoff et al. (2010) and Horstmann and Bischoff (2010a). Please note that the statistics of different lithological objects presented in Tables 1 and 2 may not be representative for the real distribution of collected fragments.
The reduction rims surrounding primary mineral grains in ureilites are less controversial. It is believed that they formed by a secondary, late-stage reduction event, probably associated with the disruption of the parent body (e.g., Warren and Kallemeyn 1992; Goodrich et al. 2004).
Oxygen isotope compositions of ureilites fall along a line of slope approximately 1 in a δ17O–δ18O diagram. This line overlaps with the carbonaceous chondrite anhydrous mineral line (CCAM) defined by Allende calcium-aluminum-rich inclusions and C2–C3 materials (Clayton and Mayeda 1988, 1996). This unique pattern among achondrites reflects oxygen isotope heterogeneity of the ureilite precursor materials.
Among the studied rocks from the Almahata Sitta strewn field, nine samples are ultra-fine-grained ureilites and 11 or 12 are coarse-grained ureilites. One of the coarse-grained specimens may not be a real ureilitic lithology as described below. Two fragments are dominated by metal–sulfide intergrowth having ureilitic lithologies attached or enclosed. All fragments have a similar degree of weathering (Table 1): W0/1 as a maximum (Wlotzka 1993), which is certainly the effect of desert weathering.
The fragments of the fine-grained variety are mineralogically similar to those described by Herrin et al. (2009), Jenniskens et al. (2009), and Zolensky et al. (2009). Typical individual recrystallized olivine grains within these fragments are always below 30 μm, in many cases below 20 μm (Figs. 1a–c and 1f; Table 1). The fine-grained texture of fragment MS-168 is shown in Fig. 1f. Low-Ca pyroxene often occurs intergrown with olivine, Ca-pyroxene, and troilite. Ca-pyroxene is mostly an interstitial phase, and FeS and kamacite occur as grains of variable grain size. Although the bulk fragments are similar in texture to one other, they vary in mineral composition from fragment to fragment (Table 1). Some have olivine core composition of approximately Fa12–16 (e.g., MS-28, MS-61, MS-154), whereas others can have olivine core compositions between ∼Fa8 (e.g., MS-20, MS-168) and ∼Fa18–21 (e.g., MS-124, MS-152, MS-161; Fig. 1a). One fragment (MS-152), mainly having olivine of Fa18–21, includes a highly reduced clast with approximately Fa1 olivines (Fig. 1b). In another case, a fine-grained ureilitic fragment contains a fragment with a significantly larger grain size compared to the surroundings (MS-124; Fig. 1c). The cores of the olivine within the coarser-grained fragment (∼Fa18–20) have a similar composition to those of the fine-grained lithology (∼Fa19–21).
Fragment MS-165 is a niningerite-bearing ureilite (Fig. 1d). Within this fragment, which is dominated by fine-grained ureilitic lithologies, an area was identified, which contains abundant niningerite and metals that have compositions (Ni: ∼3.5 wt%, Si: ∼4.4 wt%, Co: ∼0.3 wt%) similar to those in EH chondrites. As additional phases low-Ca pyroxene (up to En99), Cr-bearing troilite, and a SiO2-phase were found. Grains of SiO2 in ureilites were previously reported to occur within the reduced olivine rims (e.g., Weber et al. 2003). Some of the fine-grained ureilitic fragments contain melted areas as indicated by the occurrence of metal and metal/sulfide spherules and the presence of minerals like keilite that occur in enstatite chondrite impact melt rocks and breccias (Keil 2007). A typical specimen is fragment MS-20 containing metal-rich areas having opaques similar to those found in enstatite chondrites: niningerite, keilite, troilite, and metals with about 3.5 wt% Si and about 5.5 wt% Ni (Fig. 1e).
Based on texture and mineral compositions, the coarse-grained ureilites belong to at least six different lithologies of the parent body (Fig. 2). They all have typical grain sizes in excess of 100–300 μm. Some are very coarse grained with olivine grains up to several mm in size (Table 1). Table 1 also contains the compositions of olivine and pyroxene of the samples as well as the degrees of shock and weathering. Based on the shock classification schemes for ordinary and enstatite chondrites (Stöffler et al. 1991; Rubin et al. 1997), all coarse-grained ureilitic fragments are very weakly (S2) or weakly (S3) shocked as indicated by either undulatory extinction or the presence of planar fractures in olivine (Table 1).
Some fragments represent typical ureilites: e.g., MS-160 (Fig. 2a), MS-162, MS-157, MS-167 (Fig. 2b), MS-170. The millimeter-sized olivines within these ureilites have cores of approximately Fa12–22 and reduced rims (down to ∼Fa1; Table 1). The four fragments MS-16, MS-169 (Fig. 3a), MS-173, and MS-175 (Figs. 2c and 2d) have abundant pyroxene and the olivines in these fragments have cores with lower fayalite contents (∼Fa5–∼Fa11–14) than those within the typical coarse-grained, pyroxene-poor variety (Table 1). In fragment MS-16, pyroxene is by far the dominant phase (>80 vol%; Fig. 3b). Based on the mineralogy, it is not certain whether this rock represents an ureilitic lithology or not. The low-Ca pyroxenes are quite uniform in composition (∼Fs5–6Wo3–4.5), whereas the olivines have cores with approximately Fa5–6, but show some zoning (Table 1). The metals within the samples have about 4–5 wt% Ni. However, based on oxygen isotope composition MS-16 falls within the ureilite field (see below).
Fragment MS-156 represents a unique type of ureilitic lithology. The huge (>1 mm) olivine grains (Fig. 3c) have a remarkable internal texture with zoned olivines and a fine-grained network of FeS and metals with variable Ni-concentrations (Fig. 3d). The cores of the olivines are approximately Fa19, whereas the olivine close to the sulfide/metal assemblage is Fa<10 (Fig. 3d). The Almahata Sitta fragments MS-153 (Fig. 3e) and MS-171 (Fig. 3f) are distinctly smaller grained (typical grain size: 100–300 μm; Table 1) than the rest of the coarse-grained ureilitic lithologies. Both fragments have specific, unique textures: MS-153 contains beside olivine (∼Fa10–12) abundant pyroxene, whereas the olivine grains (∼Fa15–17) in MS-171 show a lineation (Fig. 3f).
Two meteorite fragments are dominated by metals and sulfides. In one case, such a sulfide-metal assemblage (MS-158; Fig. 4a) has an area of fine-grained ureilitic lithology attached (Fig. 4b). Highly reduced olivine is found in silicate inclusions within the sulfide-metal intergrowth. In a second fragment, the sulfide-metal assemblage is very heterogeneous and porous (MS-166). One part is sulfide-rich (troilite) enclosing coarse-grained and fine-grained ureilitic olivine (Fig. 4c). The metal has about 17 wt% Ni. In another area of the fragment, Si-poor metal with variable Ni-concentrations (about 7–20 wt%) dominates (Fig. 4d). These metals are surrounded by small Ni-rich metal grains (∼30 wt% Ni) and enclosed in a complex intergrowth of small metals, Ni-bearing sulfides (probably troilites), and Fe-oxides (Fig. 4d).
Of the 17 chondritic fragments identified, 14 belong to different enstatite chondrite groups and two to the ordinary chondrite group (Table 2). Among these chondritic varieties is also a “unique” chondrite—different from typical ordinary, carbonaceous, and Rumuruti (R) chondrites—which will be described in detail by Horstmann et al. (2010). All fragments have a similar degree of weathering (Table 2). Due to differences in texture, mineralogy, and mineral compositions, the 14 enstatite chondrite fragments represent at least six or seven different enstatite chondrites (Table 2): EH3 (MS-14), EL3/4 (MS-17, MS-164), EL6 (MS-52, MS-79), EL breccias (MS-7, MS-D, MS-174), EL impact melt rocks or impact melt breccias (MS-150, MS-159, MS-172), and EH impact melt rocks and breccias (MS-13, MS-155, MS-163). Some of the EH impact melt rocks and breccias are shock-darkened to various degree and may indicate two distinct lithologies. The degree of shock metamorphism of all chondritic samples was determined based on the classification schemes of Stöffler et al. (1991) and Rubin et al. (1997). All samples are very weakly (S2) or weakly (S3) shocked (Table 2).
The two fragments of the H-group ordinary chondrites are different from one another in mineralogy and texture: The H5 (MS-151) and H5/6 (MS-11; Fig. 5a) chondrites have different compositions of olivine and pyroxene (∼Fs17.5/∼Fa20.5 and ∼Fs14/∼Fa16.5, respectively). The fragment MS-151 is shock-darkened and the texture is barely visible in transmitted light. A higher petrologic type than H5 cannot be ruled out for this fragment. MS-11 contains shock-melted areas as indicated by the occurrence of metal and metal/troilite spherules and opaque assemblages of fine-grained metal–troilite intergrowth.
The “unique” chondrite fragment (MS-CH) is a type 3.8 ± 0.1 chondrite with a chondrule/matrix ratio of about 1.5 (Fig. 5b). Olivine is mainly Fa35–37. As the rock has a considerable abundance of mainly Ni-rich metal (Ni: ∼40 wt%, Co: ∼2 wt%), a relationship to CK chondrites (e.g., Kallemeyn et al. 1991; Geiger et al. 1993; Geiger and Bischoff 1995) can be ruled out. Based on the mineral chemistry of the silicates, the MS-CH sample has more similarities to R chondrites than to any other chondrite group (e.g., Bischoff et al. 1994; Schulze et al. 1994), but the significant abundance of metal, the lack of NiO in olivine, the lack of PGE-rich phases, and the low TiO2 concentration of the Cr-spinels demonstrate distinct differences with R chondrites. For details, see Horstmann et al. (2010).
In the following, some of the enstatite chondrite fragments will be characterized in more detail. The basic characteristics of the other fragments can be taken from Table 2. The Si-concentrations of metals in the EH and EL chondrites are distinctly different (Brearley and Jones 1998). Within the EL group, the Si content of kamacite increases from roughly 0.4 wt% in EL3s to 0.9–1.8 wt% in EL5s and finally 1.1–1.7 wt% in EL6s. Within the EH group the Si content of kamacite is about 2 wt% in EH3s, increasing to 2.6–3.5 wt% in EH4s, and roughly 4 wt% in EH6 chondrites (e.g., Keil 1968; Sears et al. 1982; Brearley and Jones 1998; and references therein). These criteria in combination with sulfide mineralogy and textural aspects were used to subdivide the Almahata Sitta E chondrite fragments.
Almahata Sitta fragment MS-14 is a very unequilibrated EH3 chondrite (Fig. 5c), has highly variable compositions of low-Ca pyroxene (Fs0.2–13; mean Fs3±4), and contains minor forsteritic olivine. Preliminary data show that the abundant metals have mean Si-, Co-, and Ni-concentrations of approximately 3.1, 0.7, and 4.2 wt%, respectively. The Ni-concentrations vary from ∼3 to ∼7 wt%. Other phases include plagioclase, a SiO2-phase, perryite, schreibersite, troilite, daubreelite, niningerite, and oldhamite.
Fragment MS-17 is an EL3/4 chondrite (Fig. 5d) having abundant chondrules. In some of these chondrules, minor forsteritic olivine occurs. The enstatites contain low, but variable Fe-concentrations (Fs0–1). Preliminary analyses show that the metals have Si-, Co-, and Ni-concentrations of about 0.5, 0.7, and 6.3 wt%, respectively. Other phases include Ca-pyroxene, plagioclase, a SiO2-phase, graphite, troilite, oldhamite, and alabandite.
Some E chondrite fragments are shock-darkened. Fragment MS-13 is an EH chondrite (Fig. 5e), in which most enstatites have small, but significant Fs-contents (up to 1.6 mole%). Based on the texture, it is probably an impact melt rock (Fig. 5e). The metals have Si-, Co-, and Ni-concentrations of roughly 2.7, 0.7, and 5.5 wt%, respectively. Other phases include plagioclase, a SiO2-rich phase, troilite, niningerite, oldhamite, and graphite.
Two samples are strongly metamorphosed EL6 chondrites (Fig. 5f). Three fragments are classified as EL chondrite breccias (see fig. 1 in Horstmann and Bischoff 2010a). The presence of keilite indicates that at least some fragments of the breccias are of impact melt origin (Keil 2007; cf. Table 2). The breccia MS-D is a highly recrystallized enstatite-rich rock which contains several clasts up to 5 mm in apparent size. Some of these clasts contain remarkably high abundances of Ca-pyroxene. The metals in MS-D have Si-, Co-, and Ni-concentrations of approximately 0.9, 0.5, and 5.5 wt%, respectively. Other yet identified phases include plagioclase, troilite, oldhamite, Zn-bearing alabandite, and keilite. Based on the texture and the presence of keilite, six impact melt rocks or impact melt breccias (three EH and three EL; Table 2) are among the enstatite chondrite samples. Some mineralogical information on these fragments is given in Table 2.
Oxygen Isotope Composition
The oxygen isotope compositions of 14 fragments (six chondritic and eight ureilitic) were determined (Table 5). All ureilite samples (MS-16, -20, -61, -124, -168, -169, -170, and -175) fall on the CCAM line (Fig. 6). Three enstatite chondrite fragments (MS-52, MS-79, and MS-D) fall within uncertainty on the TFL in the enstatite chondrite field. Samples MS-11 and MS-151 were classified as H chondrites. Considering the error limits of the oxygen isotope compositions (δ18O = ±0.2‰) these samples are H chondrites, but a slight tendency toward L chondrites is obvious (Fig. 6). The oxygen isotope composition of fragment MS-CH plots at the low δ17O border of the R-chondrite field and will be discussed in Horstmann et al. (2010).
Table 5. Oxygen isotope composition of Almahata Sitta fragments. Data are reported in ‰ relative to standard mean ocean water.
The measured activity concentrations for the detected cosmogenic radionuclides (22Na, 54Mn, 46Sc, 26Al, 57Co, and 60Co) are given in Table 3. The detection of 46Sc (half-life: 83.8 days) in MS-CH, and of 54Mn (half-life: 312.2 days) and 57Co (half-life: 271.8 days) in both samples clearly indicates that these fragments result from a very recent meteorite fall consistent with the Almahata Sitta event. In particular, the value for 46Sc in MS-CH clearly indicates that this fragment results from a very recent meteorite fall, about 335–420 days prior to the measurement. For the sample MS-D, one can give only a much weaker estimate of the date of fall, using the 57Co and the 54Mn data. The range is 300–600 days based on the available data in the literature for the measured activity concentrations for both radionuclides in chondrites (e.g., Evans et al. 1982). All cosmogenic radioisotopes except 60Co agree within one sigma in the two analyzed fragments. This particular difference can therefore hardly be explained by different shielding positions of the fragments within the same parent body, as more than one radioisotope should be affected. A more plausible explanation seems to be a local inhomogeneous distribution of Co and Ni itself, the most important target materials for the production of this isotope. It could vary up to a factor of 2 with respect to the average values reported for both types in literature. Either there is more Ni and Co in MS-D than in MS-CH or less Ni and Co in MS-CH with respect to MS-D. The sample masses are rather small and such inhomogeneous distributions start to be important even for gamma spectrometric measurements. Usually, sample masses from 50 g upward are measured, and hence this effect is averaged out over the whole sample. For the different values in 60Co, further analysis is needed to interpret the data correctly. The long-lived spallation product 26Al has reached its saturation activity. The average production rate is well in agreement with the values cited in literature (Bhandari et al. 1993).
The concentrations of the natural radionuclides 232Th and 238U as well as for Knat in the meteorite specimens are listed in Table 4. They are well in accordance with the values reported in literature (Wasson and Kallemeyn 1988). Inhomogeneous distribution could also explain why the value for uranium differs by more than one sigma, a significant difference, between the two measured samples.
A Common Asteroidal Origin of Chondritic and Ureilitic Lithologies in Almahata Sitta
Although we have short-lived cosmogenic radioisotope data on only two samples, and we do not have oxygen isotope data on all samples, we believe that most, if not all different lithologies described within this study belong to the Almahata Sitta meteorite fall. In the following, we will discuss the “pros” and “cons” of this hypothesis in detail. The main “pro” arguments can be summarized as follows:
1The detection of the short- and medium short-lived cosmogenic nuclides 46Sc, 57Co, and 54Mn clearly indicates that the chondritic fragments MS-CH and MS-D result from a fresh meteorite fall consistent with the Almahata Sitta event in October 2008. In this respect, it is important to mention that these two samples are mineralogically completely different from each other and also different from the main ureilitic lithologies from the Almahata Sitta strewn field.
2No sample has a weathering degree higher than W0/1 as indicated by the only slight brownish taint in transmitted light and/or by only very thin rinds of weathering products surrounding the metal grains (Tables 1 and 2). Thus, all samples have a very similar degree of weathering.
3Although most small fragments from the strewn field appear to represent fragments of a single lithology, preliminary studies show that at least some fragments contain two different lithologies (e.g., MS-124, MS-152, MS-158, MS-166; see above).
4Among the fragments at least seven different E chondrite lithologies were detected. Considering the actual meteorite flux, about 75% of meteorite falls (more than 90% of the chondrite falls) represent ordinary chondrites. Enstatite chondrites are relatively rare (below 2% of the actual meteorite flux) and such a high number of fresh, different E chondrite meteorite falls in just one small area can only be explained with a common origin in the asteroid 2008 TC3. Even if we consider EL and EH chondrite breccias in both cases, we would need at least two different E chondrite fall events.
5Meteorite falls are usually eyewitnessed by many local people. For the many different rock types found within the Almahata Sitta strewn field, a considerable number of recent meteorite falls (at least six!) is required. However, such eyewitness reports do not exist.
6The discovery of several new unique meteorite fragments (having so far unknown textures and mineralogy; e.g., MS-CH, MS-166, and MS-158) in a small area is best explained with a breakup of a polymict asteroid.
Nevertheless, it remains to demonstrate definitively that all fragments from the Almahata Sitta strewn field were components of asteroid 2008 TC3. Because of the lack of short-lived cosmogenic radioisotope data on all studied fragments there is still some uncertainty. As more than 90% of chondrite falls are ordinary chondrites, considering fall statistics at least for the ordinary chondrite group a certain chance of a recent fall exists. We mentioned above that meteorite falls are usually seen by many eyewitnesses. This may not be the case in the relatively uninhabitated desert area of Sudan.
Considering all “pros” and “cons,” and available data, we suggest that it is most likely that all the different chondritic and achondritic components so far found in the Almahata Sitta strewn field have a common origin in asteroid 2008 TC3.
The ureilite fragments fall in the ureilite field in a 3-oxygen isotope diagram (Fig. 6). They do not, however, show any systematic relationship between Δ17O and Fa-content of the olivine (cf. Tables 1 and 5), as has been described for other ureilites (e.g., Mittlefehldt et al. 1998). The fine-grained ureilites cluster at high Δ17O, whereas the coarse-grained ureilite fragments are more 16O-rich (Fig. 6). The E chondrite fragments all fall within error on the TFL in the E chondrite field (Fig. 6) supporting the petrological classification. The H chondrites fall within the ordinary chondrite field. The oxygen isotopes of fragment MS-CH (see Horstmann et al. 2010) are closely linked to R chondrites. The result illustrates that the Almahata Sitta breccia is fragmented on a very small scale.
Mixing of Different Rock Types in Asteroids—Fragments in Chondrites
In ordinary chondrite breccias, fragments from other ordinary chondrite groups are very rare and are summarized in Bischoff et al. (2006). These include: (1) intensely shocked H-group chondrite fragments in the LL chondrite St. Mesmin (Dodd 1974); (2) an LL5 clast in the Dimmitt H chondrite regolith breccia (Rubin et al. 1983); (3) an L-group melt rock fragment in the LL chondrite Paragould (Fodor and Keil 1978); (4) fragments in Adzhi-Bogdo (LL3–6), which appear to derive from L-group chondrites (Bischoff et al. 1993, 1996); (5) a troctolitic clast with an H chondrite oxygen isotopic composition in the Yamato-794046 (L6) chondrite (Prinz et al. 1984); (6) an L chondritic inclusion in the Fayetteville H chondrite regolith breccia (Wieler et al. 1989); and (7) a fragment of H chondrite parentage within the Ngawi LL chondrite (Fodor and Keil 1975).
CM-type fragments occur in different groups of chondrites: A small CM chondrite clast with a matrix of phyllosilicates and sulfides was observed in the Magombedze (H3–5) chondrite breccia (MacPherson et al. 1993). In addition, carbonaceous clasts were described to occur in the H-group ordinary chondrite breccia Dimmitt (Rubin et al. 1983) and the H-group chondrites Abbott and Plainview (Rubin and Bottke 2009). Other possible carbonaceous clasts in various ordinary chondrites are given by Keil (1982).
Bischoff et al. (2006) concluded that “asteroids are generally modified by two kinds of hypervelocity impacts: frequent impacts that crater the surface, and large rare impacts that damage the whole asteroid and create large volumes of rubble.” A major fraction of meteorite breccias is created by these large impacts (e.g., Scott 2002; Scott and Wilson 2005). Meteorites containing foreign clasts are typically regolith breccias, but the absence of solar wind gases excludes the possibility that Almahata Sitta is a regolith breccia (Ott et al. 2010). As shown in this study, Almahata Sitta has at least 10 different ureilitic lithologies and at least another 10 different—in one case “unique” (Horstmann et al. 2010)—chondritic rock types. Based on our own and available data, most of the Almahata Sitta fragments appear to be of ureilitic origin supporting the classification of the meteorite as a polymict ureilite. In fact, no other “normal” polymict ureilite has such a high abundance of exotic clasts. Considering the 10 ureilitic lithologies and their extraordinary diversity alone, it is clear that the coexistence of these lithologies in Almahata Sitta can only be the result of a gigantic and catastrophic disruption and breakup of the ureilite parent body delivering and producing all these different ureilitic lithologies (Fig. 7). Goodrich et al. (2004) discuss the scenario that the delivered material possibly reaccreted into second-generation asteroids of mixed ureilitic material, and that these offspring bodies are sampled by the current ureilite collection. We suggest that at the time of accretion of second-generation asteroid(s), all sorts of chondritic fragments may have been present in a debris disk around the Sun. The chondritic and ureilitic components were mixed and accreted to a second-generation asteroid parental to asteroid 2008 TC3. Clearly, 2008 TC3 is part of a second-generation asteroid. From the study of Almahata Sitta fragments, we have no information about the presence of primordial dust in the region of reaccretion. As most of the studied fragments consist of a single lithology, we suggest that the highly porous, fine-grained ureilite material described in Almahata Sitta by Jenniskens et al. (2009) may represent only the weak and relatively unconsolidated matrix (probably of low modal abundance) which surrounded the monolithic clasts. Our studied small fragments—especially the chondritic ones—are not porous and are presumably not inherently weak. Considering the Almahata Sitta bulk rock in general, we have no information on the strength and lithological connections of the various chondritic and achondritic components in Almahata Sitta. Therefore, Almahata Sitta is very different from other polymict ureilites: In other polymict samples, ureilitic and exotic (chondritic) components can be found well-consolidated in one thin section. In addition, other polymict ureilites do not contain such an apparent abundance of “exotic” clasts, as it is the case for Almahata Sitta. Enstatite chondrites and unusual fragments with affinities to R chondrites have never been reported to occur as clasts in polymict ureilites. We suggest that Almahata Sitta was much more loosely lithified and porous than Kaidun (Zolensky and Ivanov 2003), which is a well-consolidated breccia, in which the relationship between different lithologies can be studied in much more detail than in the case of Almahata Sitta. Asteroid 2008 TC3 probably consisted of centimeter-sized(?), loosely agglomerated components that broke up mainly into monolithic fragments along their original boundaries during breakup in the atmosphere (Fig. 7) (Horstmann and Bischoff 2010b).
Kaidun—Probably the Closest Analog to Almahata Sitta
The Kaidun breccia basically consists of chondritic components and not of achondritic constituents. But, based on the huge variety of different types of chondritic fragments, we suggest that Kaidun is the closest known analog to Almahata Sitta. Based on the exceptional variety of rock types, Zolensky and Ivanov (2003) characterized the Kaidun microbreccia as a “harvest from the inner and outer asteroid belt.” Kaidun consists almost entirely of millimeter- and sub-millimeter-sized fragments of EH3–5, EL3, CV3, CM1–2, and R chondrites (Ivanov 1989; Ivanov et al. 2003; Zolensky and Ivanov 2003, and references therein), contains C1 and C2 lithologies, fragments of impact melt products, new enstatite-bearing clasts, phosphide-bearing fragments, clasts of Ca-rich achondrite, possibly aubritic materials, and alkaline-enriched clasts (Ivanov 1989; Ivanov et al. 2003; Zolensky and Ivanov 2003; Kurat et al. 2004). The alkaline-enriched clasts are similar to the granitoidal clasts found in the Adzhi-Bogdo ordinary chondrite regolith breccia (Bischoff et al. 1993, 1996; Sokol and Bischoff 2006; Sokol et al. 2007a, 2007b; Terada and Bischoff 2009). In addition, a possible ordinary chondrite clast has been characterized by Mikouchi et al. (2005). Thus, after the polymict breccia Kaidun (Zolensky and Ivanov 2003) Almahata Sitta is a new extraordinary breccia for future studies. The coexistence of many different types of clasts in Kaidun supports our hypothesis that most, if not all, studied fragments derive from asteroid 2008 TC3.
The Mismatch in Reflectance Spectra
According to Gaffey et al. (1993), ureilites should be derived from S-type asteroids. These authors distinguish between pyroxene-poor ureilites (class S(I)), clinopyroxene-bearing ureilites (class S(II)), clinopyroxene and orthopyroxene-bearing ureilites (class S(III)), and orthopyroxene-bearing ureilites (class S(IV)). However, the reflectance spectra of asteroid 2008 TC3 are similar to those of B- and F-type asteroids (Jenniskens et al. 2009). Based on Gaffey et al. (1993), type B asteroids should contain iron-poor hydrated silicates and type F asteroids should have hydrated silicates and organics. Jenniskens et al. (2009) did not find hydrated silicates in the Almahata Sitta meteorite sample, but came to the somewhat surprising conclusion that Almahata Sitta is similar to a type F asteroid. The asteroid 2008 TC3 analyzed in space was a mixture of very different lithologies. It is not certain that all its different lithologies have been recognized yet. They probably have not. Perhaps lithologies with hydrated silicates also existed within asteroid 2008 TC3. Based on the findings presented in this study, the reflectance spectrum of asteroid 2008 TC3 should be re-evaluated.
Acknowledgments— We thank U. Heitmann (Münster) for sample preparation. N. Albrecht (Göttingen) is thanked for the preparation of the oxygen isotope analyses, and Ed Scott (University of Hawaii) for interesting discussions on the possible accretion scenario of asteroid 2008 TC3. We also thank the referees Klaus Keil, Alan Rubin, and Harold Connolly for constructive reviews, and the associate editor Cyrena Goodrich for thoughtful comments and discussions.