Thermal and fragmentation history of ureilitic asteroids: Insights from the Almahata Sitta fall


Corresponding author. E-mail:


Abstract– The Almahata Sitta fall event provides a unique opportunity to gain insight into the nature of ureilitic objects in space and the delivery of ureilite meteorites to Earth. From thermal events recorded in the mineralogy, petrology, and chemistry of ureilites recovered from the fall area, we reconstruct a timeline of events that led to their genesis. This history is similar to that of other known ureilites and supportive of a disrupted ureilite parent body hypothesis. Temperatures of final mantle equilibrium were 1200–1300 °C, but this high-temperature history was abruptly terminated by rapid cooling and reduction associated with pressure loss. The onset of late reduction reactions and onset of rapid cooling must have been essentially simultaneous, most likely engendered by the same event. Cooling rates of 0.05–2 °C h−1 determined from reversely zoned olivines and pyroxenes in Almahata Sitta imply rapid disassembly into fragments tens meters in size or smaller. This phenomenon seems to have affected all known portions of the ureilite parent body mantle, implying an event of global significance rather than localized unroofing. Reaccretion of one or more daughter asteroids occurred only after significant heat loss at minimum time scales of weeks to months, during which time the debris cloud surrounding the disrupted parent was inefficient at retaining heat. Fragments initially dislodged from the ureilite parent body mantle underwent subsequent size reduction and mixed with various chondritic bodies, giving rise to polylithologic aggregate objects such as asteroid 2008 TC3.


Asteroid 2008 TC3, the parent object of the Almahata Sitta fall, was discovered on an Earth-intersecting trajectory just prior to entry into Earth’s atmosphere (Jenniskens et al. 2009). Recovery of meteorite fragments from the fall, the majority of which are ureilites, has provided a unique opportunity to compare remotely sensed spectral data with laboratory determined mineralogy and composition. The event has also provided insight into the nature of ureilitic objects in space and the delivery of ureilite meteorites to the Earth’s surface.

Other works in this volume describe asteroid 2008 TC3 as a loosely consolidated polylithologic aggregate based on the high altitude at which it disintegrated in the Earth’s atmosphere, the amount of mass lost, and the recovery of different meteorite types within the strewn field (Jenniskens et al. 2009; Shaddad et al. 2010; Zolensky et al. 2010). Such an object seems to contrast with what is known of previous recorded ureilite falls, none of which are polymict (Table 1) and none have known association with other meteorite types (Grady 2000). Polymict ureilites do however make up a significant fraction of finds (e.g., Goodrich et al. 2004). Recovered specimens of Almahata Sitta tend to be small, most <10 g, and individual specimens are typically mostly or wholly fusion crusted so the contact relationship between different specimens is unclear from those that we examined.

Table 1.   Recovered mass and classification of all six observed ureilite falls (Grady 2000).
FallRecovered mass (g)Classification
Novo-Urei (Russia, 1886)1900Monomict/unbrecciated
Dyalpur (India, 1872)300Monomict/unbrecciated
Lahrauli (India, 1955)>900Monomict/unbrecciated
Haverö (Finland, 1971)1544Monomict/unbrecciated
Jalanash (Mongolia, 1990)700Monomict/unbrecciated
Almahata Sitta (Sudan, 2008)10,500Polymict

Ureilites are carbon-rich ultramafic achondrites consisting predominantly of olivine and low-Ca pyroxene, most commonly pigeonite (e.g., Goodrich 1992; Mittlefehldt et al. 1998). Orthopyroxene and augite are found in many ureilites. Common accessory phases include iron metal, iron sulfide, graphite, diamond, and londsdaleite. Aluminous phases such as plagioclase are absent. Like other ureilites, Almahata Sitta contains microdiamonds (Steele et al. 2009; Ross et al. Forthcoming) and reverse-zoned rims on olivine grains. We examined the record of thermal processes preserved in ureilitic specimens of Almahata Sitta using mineral thermometry combined with chemical and thermal diffusion modeling, and compared these results with other ureilites to determine if any identifiable feature might account for the polylithologic and “rubble pile” character of asteroid 2008 TC3. In this work, we also present a summary of what is known of the thermal history of ureilites and what that implies for their genesis. We place this information in the context of the Almahata Sitta fall in an effort to gain new insight into both the genetic history of ureilites, and the architecture and formation of modern ureilitic asteroids.

Samples and Analytical Methods

Some 600 individual stones of Almahata Sitta have been recovered from the Nubian Desert strewn field through efforts of the University of Khartoum and the Almahata Sitta Consortium (Shaddad et al. 2010). Samples from several meteorites were made available for research and mineralogic descriptions of many of these samples are presented by Zolensky et al. (2010). Four of these samples were found to be chondrites (H5, L4, EH6, and EL6) collected from the Almahata Sitta strewn field, as described by Zolensky et al. (2010) and Shaddad et al. (2010). Bischoff et al. (2010) also report a variety of chondritic materials in Almahata Sitta samples. In this work, we rely solely on samples sourced to the Almahata Sitta fall event by mass and location in the strewn field by Shaddad et al. (2010) for our interpretations of the asteroid 2008 TC3 and the Almahatta Sitta fall. We present analytical work performed on nine ureilitic samples from the strewn field, namely samples 7, 15, 29, 32, 33, 39, 51, 53, and S138.

A polished surface of each sample was created and examined first by scanning electron microscopy (SEM) employing energy dispersive spectroscopy (EDS) to get an overview of the mineralogy and textures of the samples and then by electron probe microanalysis (EPMA) to obtain quantitative wavelength dispersive spectroscopy (WDS) analyses. These electron beam techniques were performed at NASA Johnson Space Center using a JEOL 6910LV SEM and Cameca SX100 electron microprobe. A variety of natural and synthetic mineral standards were used for external calibration of WDS measurements performed at typical operating conditions of accelerating voltage = 15 kV and beam current = 20 nA. Semiquantitative EDS analyses to examine narrow zoning profiles were performed under similar operating conditions. Analyses of olivine and pigeonite grain margins required the maximum spatial resolution afforded by our instruments and technique. Care was taken to avoid the incorporation of multiple Mg or Fe-bearing phases within the interaction volume of each analysis. In addition to Almahata Sitta, the following nonpolymict ureilites from the U.S. Antarctic Meteorite Collection were also examined for comparison using methodology identical to that described above: Allan Hills (ALH) A77257, Cumulus Hills (CMS) 04048, Elephant Moraine (EET) 87511, EET 87517, EET 90019, EET 96042, Pecora Escarpment (PCA) 82506, and the augite-bearing ureilites ALH 84136, EET 96293, Lewis Cliff (LEW) 88201, and Meteorite Hills (MET) A78008.


Mean mineral core compositions for individual samples were determined through multiple analyses, typically of several grains in each sample. Mineral compositions used for thermometry are presented in Table 2. Differences in core mg# (mole% Mg/[Mg + Fe]) compositions of ferromagnesian silicate phases between samples are apparent but within each ureilitic clast core compositions are nearly uniform with pyroxenes slightly more magnesian than olivine. Comparison of these data with other ureilites can be found in Zolensky et al. (2010).

Table 2.   Mean compositions of pyroxene cores from Almahata Sitta and Antarctic ureilites determined by EPMA and reported as weight percent oxides. These mean compositions were used to estimate pigeonite equilibrium temperatures of Singletary and Grove (2003) (S&G) for low calcium pyroxene-olivine type (LCP-ol) samples and also two-pyroxene equilibrium temperatures of Kretz (1982) and Brey and Köhler (1990) (B&K; pressure = 1 kbar) for augite-bearing phase assemblages (LCP-ol-aug). Thermometer errors are reported. Low-calcium pyroxene (LCP) is used ambiguously to refer to either orthopyroxene or pigeonite.
MeteoriteSample#MineralogyPhaseNa2OMgOAl2O3SiO2CaOTiO2Cr2O3FeOMnOTotalmg#T (°C) (S&G)T (°C) (Kretz)T (°C) (B&K)
Almahata Sitta7LCP-ol-augLCP0.0431.651.0456.122.720. 1250 ± 601213 ± 81
Almahata Sitta15LCP-ol-augLCP0.0431.660.9456.342.480.131.027.390.57100.688 1224 ± 601186 ± 79
Almahata Sitta29LCP-olLCP0.0925.601.1754.345.570.111.3810.990.4499.7811232 ± 83  
Almahata Sitta32LCP-olLCP0.0432.060.8457.172.620.130.815.470.4899.6911304 ± 82  
Almahata Sitta33LCP-olLCP0.1128.850.9854.217.510.250.716.410.5099.5891264 ± 90  
Almahata Sitta39LCP-olLCP0.0527.130.6455.063.550.061.2411.960.43100.1801240 ± 78  
Almahata Sitta51LCP-olLCP0.0730.570.7756.264.740.130.845.600.4899.4911288 ± 86  
Almahata Sitta53LCP-olLCP0.0527.060.6054.473.430.051.2112.200.4399.5801239 ± 77  
Almahata SittaS138LCP-olLCP0.0628.730.8855.454.690.091.118.370.4599.8861264 ± 83  
ALH 84136 LCP-ol-augLCP0.0534.920.5056.942.620.120.933.060.5199.695 1277 ± 601253 ± 84
EET 96293 LCP-ol-augLCP0.0331.311.2854.402.500.131.137.900.5299.288 1236 ± 601185 ± 79
LEW 88201 LCP-ol-augLCP0.0433.160.8555.832.680.120.935.950.45100.091 1290 ± 601241 ± 83
META78008 LCP-ol-augLCP0.1227.382.3452.952.300.131.3212.620.4099.679 1300 ± 601169 ± 78
ALHA77257 LCP-olLCP0.0429.970.6055.463.310. ± 81  
CMS 04048 LCP-olLCP0.0426.440.6553.843.510.051.1812.500.3998.6791235 ± 77  
EET 87511 LCP-olLCP0.0330.521.1854.412.350.071.328.610.4498.9861279 ± 79  
EET 87517 LCP-olLCP0.0433.380.5356.962.550.140.595.360.42100.0921307 ± 82  
EET 90019 LCP-olLCP0.0630.030.8655.254.870.141.206.640.4999.5891279 ± 85  
EET 96042 LCP-olLCP0.0827.040.7354.784.630.081.4110.020.4599.2831248 ± 82  
PCA 82506 LCP-olLCP0.0727.550.7054.422.950.051.1911.750.4499.1811246 ± 77  

Two of the Almahata Sitta samples, sample 7 and sample 15, contain augite. Augite is a major component of sample 15, occupying nearly 40 modal percent in a polished section, unusually high for a ureilite. This specimen is coarse grained (1–2 mm) relative to the size of the section (10 mm2) so we cannot be certain it is representative of the entire rock mass, but textural or mineralogic variations were not apparent in the portion of hand specimen from which fusion crust had broken away. In sample 7 augite occurs only as an accessory phase. We applied the two-pyroxene thermometers of Kretz (1982) and Brey and Köhler (1990) to mean pyroxene core compositions from these two augite-bearing ureilitic samples to arrive at pyroxene equilibrium temperatures of 1190–1250 °C (Table 2; Fig. 1). These temperatures overlap with pyroxene equilibrium temperatures from the Antarctic augite-bearing ureilites that we also analyzed.

Figure 1.

 Pigeonite equilibrium temperatures of Singletary and Grove (2003) (diamonds) are presented along with two-pyroxene equilibrium temperatures of Kretz (1982) (squares) and Brey and Köhler (1990) (triangles) from mineral cores in Almahata Sitta and Antarctic ureilites.

Olivine and low-calcium pyroxene (LCP), a term used ambiguously refer to orthopyroxene or pigeonite, were present in all of the ureilitic specimens that we examined. The LCP phase in ureilites is more often pigeonite than orthopyroxene, indicative of high final equilibrium temperatures. In specimens of Almahata Sitta, LCP compositions are often consistent with orthopyroxene in that they contain Wo < 5. Structurally, however, at least some of these LCP with Wo < 5 are pigeonite (Mikouchi et al. 2010; Zolensky et al. 2010).

The olivine-pigeonite-melt thermometer of Singletary and Grove (2003) is based on the assumption that pigeonite compositions were in equilibrium with olivine and basaltic melt formed during “smelting” of the ureilite protolith. In their experiments, the composition of pigeonite was shown to be a temperature-dependent variable and equilibrium temperatures may be estimated from pigeonite compositions using the expressions that they defined. Applying this thermometer to LCP in the (non–augite-bearing) samples we examined, we arrive at temperatures of 1240–1300 °C (Table 2; Fig. 1) in agreement with the range reported for other ureilites (Singletary and Grove 2003) and also the range determined for some Antarctic ureilites in this study. Magnesian rims of olivine grains are a characteristic feature of ureilites and were observed to occur in all of the ureilitic specimens of Almahata Sitta, but in none of the chondritic specimens that we examined. Magnesian rims are irregular and sometimes extend into the interior of grains along stringers delineated by trails of metal inclusions. They often occur in contact with low-Ni metal and interstitial carbon phases. Compositional profiles across high-mg# olivine margins performed on sample 15, presented in Table 3 and Fig. 2a, reveal that these high-mg# rims are approximately 25 μm in width and account for a difference in mg# of approximately 10. Such olivine rims are typical of ureilite olivines (Berkley et al. 1980; Miyamoto et al. 1985). In a pyroxene-dominated porous sublithology of sample 7 described by Jenniskens et al. (2009) and Zolensky et al. (2010), we observe magnesian rims 4–6 μm in width on pigeonite grains in contact with low-Ni iron metal, a silica phase, and minor graphite (Table 3; Fig. 2b).

Table 3.   Examples of compositional profiles across reverse-zoned margins of olivine and pigeonite indicating changes in mg# as a function of distance from grain margin.
Profile 1 (olivine)Profile 2 (olivine)Profile 3 (pigeonite)Profile 4 (pigeonite)
Distance (μm)mg#Distance (μm)mg#Distance (μm)mg#Distance (μm)mg#
Figure 2.

 a) Top: Backscattered electron image of Almahata Sitta sample 15, a coarse-grained augite-bearing specimen. Fusion crust is visible at bottom of image. Bottom: Compositional profile across a reverse-zoned margin of olivine showing mg# as a function of distance from grain margin. b) Top: Backscattered electron image of a fine-grained, porous, pigeonite-dominated lithology in Almahata Sitta sample 7 showing fossilized reduction texture. Bottom: Compositional profile across a reverse-zoned margin of pigeonite showing mg# as a function of distance from grain margin.

We use the observed disequilibrium of reverse-zoned rims on olivine and pyroxene from Almahata Sitta to estimate minimum cooling rates using the asymptotic cooling model of Ganguly et al. (1994) and Fe-Mg interdiffusion kinetics of olivine (activation energy [Q] = 275 kJ mol−1; Chakraborty et al. 1994) and pyroxene (Q = 240 kJ mol−1; Ganguly and Tazzoli 1994; Ganguly et al. 1994) (Fig. 3). From initial temperatures of 1200–1300 °C down to 800 °C, mean cooling rates of 0.4–2 and 0.05–0.2 °C h−1 were estimated from pyroxene and olivine, respectively. The slight differences in these two estimates from different minerals could be related to differences in closure temperature, uncertainty in input parameters, and difficulty in measuring profiles precisely oblique to grain boundaries. Regardless, each of the two estimates indicates very rapid cooling during formation of the reduced rims. Such rapid cooling rates are consistent with previous estimates from other ureilites based on similar methods (Toyoda et al. 1986; Chikami et al. 1996; Goodrich et al. 2001) as well as cooling rate estimates derived from wavelengths of augite exsolution lamellae in pigeonite (Mori and Takeda 1983; Chikami et al. 1996). More importantly, our cooling rate estimates based on pigeonite rims from sample 7 are nearly identical to and overlap with rates of 0.2–5 °C h−1 determined by wavelengths of augite exsolution lamellae within this same sample (Mikouchi et al. 2010).

Figure 3.

 Cooling rate equation for pyroxene (Ganguly et al. 1994).


The History of Almahata Sitta and the Story of Ureilites

The majority of ureilitic samples of the Almahata Sitta fall are very similar to other ureilites, the principal exception being the “pyroxene-dominated, very porous, highly reduced lithology” described by Jenniskens et al. (2009) and Zolensky et al. (2010). By extension, we infer that the conditions of formation for Almahata Sitta ureilites are also similar to those of other ureilites. In this discussion, we detail discernable aspects of the genesis of ureilites in chronological order focusing on recorded thermal events and highlighting evidence from Almahata Sitta samples. This summary of what is known of ureilite genesis derives from the existing body of literature on the topic developed over the past several decades. Throughout this discussion, we also seek to evaluate which aspects of ureilite genesis might have led to formation of a loosely consolidated polylithologic aggregate object such as asteroid 2008 TC3 and what new insights the Almahata Sitta samples can provide for ureilite genesis.

A sketch outlining the history of Almahata Sitta is presented in Fig. 4. The bulk of this history is shared with other ureilites, which can be summarized as follows: A carbon-rich, chondrite-like protolith accreted to an asteroid of significant size referred to as the ureilite parent body (UPB). The UPB existed in its original form for only a brief period of time in the early solar system and is thus distinct from ureilitic asteroids that exist today, although one asteroid or family of asteroids may be a major remnant of the UPB (Jenniskens et al. 2010). Several radiogenic isotopic systems suggest that ureilites formed as ultramafic rocks early in solar system history, close to 4.55 Ga (e.g., Takahashi and Masuda 1990; Goodrich and Lugmair 1995; Torigoye-Kita et al. 1995a, 1995b; Lee et al. 2009). The UPB mantle underwent heating and partial melting, but is unlikely to have reached a predominantly molten “magma ocean” state (Clayton and Mayeda 1988; Scott et al. 1993). Ureilites are thought to be mantle residues from the UPB (Takeda 1987; Warren and Kallemeyn 1992; Scott et al. 1993; Mittlefehldt et al. 1998; Singletary and Grove 2003; Kita et al. 2004). The UPB history of ureilites ended when one or more catastrophic disruption events subjected hot mantle material to rapid cooling and pressure loss (Berkley et al. 1980; Berkley and Jones 1982; Takeda 1989; Warren and Kallemeyn 1992; Scott et al. 1993; Goodrich et al. 2004; Warren and Huber 2006). These events are fossilized in reverse-zoned rims commonly observed on ureilite olivines and less distinctly as reverse-zoned rims on ureilite pyroxenes. Debris of the disrupted UPB accreted into one or more daughter asteroids. Reaccretion was necessarily cold and there is no evidence to suggest that any portion of later-generation (post-UPB) ureilitic asteroids ever exceeded blocking temperatures for silicate minerals for any significant length of time. At some point, during post-UPB history at least some ureilitic fragments became mixed with chondritic material and were further brecciated. Descendants of the original UPB must still exist in the asteroid belt today and occasionally calve objects like asteroid 2008 TC3 toward the inner solar system (Jenniskens et al. 2010; Welten et al. 2010).

Figure 4.

 Sketch of the inferred history of Almahata Sitta. Beginning with a chondrite-like protolith [1], the UPB accreted to sufficient size to undergo metal-silicate segregation and silicate partial melting. Ureilites are the mantle restites of these processes [2]. The UPB was catastrophically disrupted [3] while still hot, and dislodged fragments reaccreted into one or more daughter asteroids, which probably experienced further collisions over the course of time [4]. In a recent collision, the object 2008 TC3 was ejected from the asteroid belt (Welten et al. 2010) [5] toward an Earth-crossing trajectory where it eventually collided with the Earth’s atmosphere breaking apart at high altitude [6]. Fragments of this burst were recovered from the Nubian Desert strewn field [7] and became available for study.

Genesis of ureilites remains a controversial and frequently reinterpreted subject. For example, Rubin (2006) presents an alternative view that departs significantly from the preceding narrative. The discussion to follow is a more detailed but somewhat thermally biased chronology divided into four stages: a protolith stage (I), a UPB stage (II), a disruption stage (III), and a stage encompassing the entire subsequent cold history (IV).


It has long been inferred that ureilites are a group of achondrites derived from a carbonaceous chondrite-like protolith (e.g., Vdovykin 1970; Clayton et al. 1976; Higouchi et al. 1976; Wasson et al. 1976; Takeda 1987; Tomeoka and Takeda 1989). The two principal lines of evidence to support this come from (1) the presence of abundant carbon thought to be primary (Berkley and Jones 1982; Goodrich and Berkley 1986) and (2) oxygen-isotope compositions that range along a δ17O–δ18O line of slope = 1, essentially coinciding with the carbonaceous chondrite anhydrous mineral line (Clayton and Mayeda 1988, 1996; Franchi et al. 1998). Despite this, ureilites cannot be directly linked to any particular group of carbonaceous chondrite and may well not be derived from anything identical to any known carbonaceous chondrites (Goodrich 1999; Qin et al. 2010). Wide variation in oxygen-isotope composition and mg# might even suggest diversity among precursor materials (Clayton and Mayeda 1988, 1996). Recent work has postulated kinship between the ureilite protolith and anhydrous, carbon-deficient chondrite types based on chemical affinities (Warren Forthcoming) as well as the roster of chondrite types observed in the Almahatta Sitta fall (Jenniskens et al. 2010; Zolensky et al. 2010). High-temperature anatexis, however, has overwritten many mineralogical, petrological, and compositional characteristics so that little of the nebular and premetamorphic history of ureilites can be directly determined with present analytical methods. Advancements in stable isotope studies promise to increase our knowledge of the ureilite protolith in coming years. A number of ureilites contain trapped noble gases in chondritic abundances and with a fractionated “planetary” type pattern such as that observed in CM chondrites (Müller and Zähringer 1969; Bogard et al. 1973; Wacker 1986; Goodrich et al. 1987; Ott et al. 1993). The extent to which this noble gas composition is nebular has been an enduring discussion (Berkley and Jones 1982; Murty et al. 2010). Given recent observations of organic compounds with limited thermal stability, such as PAHs and amino acids, in Almahata Sitta specimens (Callahan et al. 2009; Morrow et al. 2009; Sabbah et al. 2009, 2010) it seems plausible that nebular components may well have been added to ureilitic lithologies in their cold post-UPB history, perhaps arriving together with chondritic materials observed in polymict ureilites as well as the Almahata Sitta fall.

Stable UPB: Heating and Partial Melting

The simplest model for ureilite genesis involves a common large parent body, the UPB, early in the history of the solar system. This ancestral asteroid is temporally and conceptually distinct from modern ureilitic asteroids. Given that there are several known types of carbonaceous chondrites, each presumably with different nebular origins, it might seem surprising that ureilites are the only known variety of carbonaceous achondrites, and even more surprising that all should be derived from a single ancestral parent body. However, there are three primary lines of evidence to suggest that a single UPB is the most plausible scenario to explain genesis of the vast majority of ureilites or that the seemingly unique series of events required to produce ureilites must have been common within a particular region of the early solar system during a restricted period of time. First, ureilites share unique mineralogy, chemistry, and textures such as reduced rims on olivine grains, the presence of microdiamonds, and a complete lack of feldspars (except in polymict ureilites). Second, as we will discuss later in this section, capping pressures for carbon reduction reactions at relevant temperatures require at least 100 km of overburden for ureilites of low mg# (Walker and Grove 1993), implying an asteroid of >200 km in diameter, although estimates of this size do vary somewhat (Berkley and Jones 1982; Goodrich et al. 1987; Warren and Kallemeyn 1992; Wilson et al. 2008). Although it is plausible that more than one such body existed in the solar system, it is also true that one such large object would hold the potential to contribute all of the flux of ureilitic material to Earth. Ureilites are the second most abundant type of achondrite with over 260 named meteorites, but they constitute <1% of observed falls. The third line of evidence, which has come to light in recent years and is strongly supported by the Almahata Sitta fall, is that the entire range of mineral and oxygen-isotopic composition observed in ureilites can also be observed in millimeter and submillimeter clasts from individual polymict ureilites (Downes et al. 2008) and in centimeter-sized fragments of asteroid 2008 TC3 (Rumble et al. 2010; Zolensky et al. 2010). That such diversity of ureilitic materials should be collected in a single meteorite or small asteroid suggests that all ureilites could easily originate from the same ancestral parent (Downes et al. 2008).

In general, ureilites are highly depleted in incompatible elements while enriched in many elements compatible during silicate partial melting (Spitz and Goodrich 1987; Warren and Kallemeyn 1992; Goodrich 1997a). Various early works have sought to highlight cumulate or paracumulate traits of certain ureilites (Berkley et al. 1980; Goodrich et al. 1987; Warren and Kallemeyn 1989), but the overall chemical and mineralogic character of ureilites is consistent with a restitic origin (cf. Boynton et al. 1976). The consensus view is that ureilites represent a restitic lithologic suite from the UPB mantle from which basaltic magmas had been extracted (Boynton et al. 1976; Takeda 1987; Warren and Kallemeyn 1992; Scott et al. 1993; Mittlefehldt et al. 1998; Singletary and Grove 2003; Kita et al. 2004; Wilson et al. 2008). Friedrich et al. (2010) demonstrated a depletion of light rare earth elements (REE) relative to heavy REE, and negative Eu anomalies in Almahata Sitta ureilite specimens. Such REE profiles are thought to be representative of the unaltered composition of typical ureilites (Crozaz et al. 2003) and are consistent with a history of basaltic partial melt removal (e.g., Goodrich et al. 2004). Pyroxene thermometry gives us an estimate of the temperatures sustained within the UPB mantle. Figure 1 shows the range of equilibrium temperatures obtained from pyroxene cores of ureilitic samples of Almahata Sitta and other ureilites. Most are between 1200 and 1300 °C and, for the purpose of this discussion, minor differences between different thermometers and different samples are insignificant. The most relevant observation is that all temperature estimates are within the range of basaltic melt generation, indicating a high final sustained temperature of the UPB mantle. Disruption of the UPB mantle must have occurred at or near peak temperatures (e.g., Walker and Grove 1993; Singletary and Grove 2003; Warren and Huber 2006).

Temperatures of partial melting, similar to temperatures of pyroxene equilibrium in these samples, must have been sustained long enough to allow for substantial melt migration within the asteroid. The preservation of wide variation in oxygen-isotopic composition dictates that the UPB mantle was never in a highly molten state, and that partial melts must have been driven off rapidly so as not to equilibrate with large volumes of solid mantle (Warren and Kallemeyn 1992; Goodrich et al. 2004; Wilson et al. 2008). This latter notion seems to be supported by additional lines of reasoning. When the conditions of isotopic heterogeneity, high-degrees of total melt extraction, high final equilibrium temperatures, and lack of petrographic evidence for trapped melts are considered in totality, it paints a picture of efficient extraction of small melt volumes during UPB anatexis. Moreover, the “missing” basaltic component from ureilites could suggest that melts were driven from the UPB surface with force enough to exceed escape velocity (Warren and Kallemeyn 1992; Goodrich et al. 2004; Wilson et al. 2008).

The final melt volume present at time of disruption, however, is unclear because of the potential for this event to have expelled the last remaining melts. We might expect that some melt component, if present, would have been preserved by rapid quenching, especially in low-porosity materials. By contrast, unbrecciated monomict ureilites lack evidence of trapped interstitial melts, and melt inclusions are only prevalent in some augite-bearing ureilites (Goodrich et al. 2001). The occurrence of melt lithologies in polymict or dimict breccias (Ikeda et al. 2000; Ikeda and Prinz 2001; Cohen et al. 2004; Kita et al. 2004) is both volumetrically minor and ambiguous in its relationship to final melt volume. Although high final equilibrium temperatures certainly permit the presence of melt, they do not in themselves imply its presence.

In summary, peak temperatures of portions of the UPB mantle preserved in ureilites were those of basaltic melt production with the degree of partial melting typically estimated at approximately 20–30%, assuming a chondrite-like protolith (Spitz and Goodrich 1987; Warren and Kallemeyn 1992; Goodrich 1997a). Final equilibrium temperatures of UPB mantle were at or near these same temperatures, yet significant volumes of melt lithology are not preserved in ureilites. These combined observations support the assertion of efficient melt extraction and indicate that the UPB mantle was disrupted while still hot.

Disruption of Hot UPB Mantle: Pressure Loss and Rapid Cooling

A favored topic in ureilite petrology is catastrophic disruption of the UPB, which existed as a large body for only a brief period in the early solar system (Takeda 1987; Warren and Kallemeyn 1992; Warren and Huber 2006; Downes et al. 2008). Converging lines of evidence reveal that the UPB mantle experienced a significant reduction in lithostatic pressure simultaneous with massive outgassing and the onset of rapid cooling (Warren and Huber 2006). The most coherent explanation for these events is large-scale fragmentation of hot UPB mantle.

Magnesian rims observed on Almahata Sitta olivines are a characteristic feature found in all ureilites. They preserve a record of chemical disequilibrium across spatial scales of tens of micrometers. They are thought to record a short-duration reduction event resulting from sudden loss of pressure favoring reduction reactions of the type FeO + C→Fe + CO, which produce iron metal from the iron oxide component in mafic silicates (Berkley and Jones 1982; Walker and Grove 1993). Since gaseous species are a product of these reactions, they require massive volume expansion to proceed. In a largely solid asteroid mantle with sufficient latent heat to reduce olivine, sudden volume expansion would necessitate large-scale open porosity or, more likely, fragmentation of the mantle (Warren and Huber 2006).

In the anomalous pyroxene-dominated porous sublithology of Almahata Sitta sample 7 described by Jenniskens et al. (2009) and Zolensky et al. (2010), we observe magnesian rims 4–6 μm in width to occur on pigeonite grains in contact with low-Ni iron metal, a silica phase, and minor graphite and diamond (Fig. 2b). This evidence suggests that the reduction mechanism MgFeSi2O6 + C→½Mg2Si2O6 + Fe + SiO2 + CO played a central role in producing the observed textures, as all of the products and reactants are preserved as discrete phases with the exception of CO gas. Similar reduction textures in Almahata Sitta pyroxenes are described by Warren and Rubin (2010). Arguably, production of CO gas could be responsible for the rounded pores visible in Fig. 2b, but the amount of CO produced would have greatly exceeded present pore volume of the rock and some gas would necessarily have escaped the system. Reduced pyroxene rims must record volume expansion and gas escape from hot UPB mantle. Indeed, pores in sample 7 were found to be interconnected in sheets surrounding ureilitic clasts (Zolensky et al. 2010). However, some pore walls are lined with vapor phase olivine of mg# similar to olivine cores, which seem unlikely to be the direct products of the reduction reactions described above unless re-equilibration of these small vapor-deposited grains has occurred.

Preservation of fine-scale reverse-zoning requires rapid cooling so that disequilibrium over short distances is not eradicated by later diffusion or recrystallization. From these textures, we were able to estimate cooling rates in the range of 0.4–2 and 0.05–0.2 °C h−1 from pyroxene and olivine, respectively, agreeing well with previous estimates from other ureilites (Mori and Takeda 1983; Toyoda et al. 1986; Chikami et al. 1996; Goodrich et al. 2001). Left undisturbed, large asteroids should cool at time scales of millions of years. For example, the acapulcoite-lodranite parent body, another partially differentiated asteroid, is shown to have cooled at much slower rates. Pellas et al. (1997) estimated cooling rates of the Acapulco meteorite to be 100 °C Ma−1 (10−7 °C h−1) at high temperatures and slowing yet further at lower temperatures. Rapid cooling rates derived from ureilites give a first-order impression of the events that affected them at the terminus of their high-temperature history. Our cooling rate estimates based on pigeonite rims from Almahata Sitta sample 7 are nearly identical to and overlap with rates determined from augite exsolution lamellae textures within this same sample by Mikouchi et al. (2010). Agreement between these two cooling rate estimates is revealing because, while both are fundamentally governed by diffusion rates, they stem from different causes. One is a measure of the internal redistribution of constituents of pyroxene crystals that occurs simply as a result of cooling. The other is a measure of the chemical gradient resulting from a fossilized reduction reaction between pyroxene and carbon. The fact that these two cooling rates should agree indicates that there was little or no temporal difference between the onset of late reduction reactions and the onset of rapid cooling, the two must have been very nearly simultaneous.

We contend that a simple “unroofing” event cannot account for the sum of all observations, and that a high degree of fragmentation resulting in destruction of the entire portion of the UPB mantle represented by known ureilites must have occurred at a time when high internal temperatures pervaded the UPB mantle. To support this hypothesis, we point to several converging lines of evidence:

  • 1 Zoning of olivine rims is a nearly ubiquitous feature in ureilites of diverse mg# and oxygen-isotope composition, suggesting a similar history of simultaneous volume expansion and rapid cooling for widely separated portions of the UPB mantle.
  • 2 The diversity within asteroid 2008 TC3 suggests that the secondary accretion environment contained fragments dislodged from distinct portions of the UPB mantle spatially separated during high-temperature metamorphism and partial melting.
  • 3 Most ureilites have little internal porosity, so the only escape for gaseous reaction products must have been along open fractures.

In an effort to quantify the scale of fragmentation resulting from disruption of the UPB, and to compare this with the fragment size within asteroid 2008 TC3, we estimated the initial size of hot dislodged fragments of UPB mantle from petrographically determined cooling rates using the heat diffusion equation of Carslaw and Jäger (1959) applied to spherical bodies in a three-dimensional cooling model. For simplicity, we ignored the effects of surface radiation and potential internal heat production from radioactive decay and exothermic reduction reactions (Singletary and Grove 2003), both of which we consider to be negligible. We applied a range of relevant thermal diffusivities from 5.0 × 10−7 (Gupta and Sahijpal 2009) to 6.3 × 10−7 m2 s−1 (Clauser and Huenges 1995) and also considered the effect of high porosity, which might decrease thermal diffusivity to as low as 2 × 10−7 m2 s−1 (Yomogida and Matsui 1983). By this estimate, mean cooling rates inferred from ureilites would be experienced by spherical bodies with a maximum diameter of approximately 10–100 m (Fig. 5). Fragments of this size are obviously much smaller than the UPB itself. Substantial disintegration of the known UPB mantle into fragments that were, as a first-order approximation, tens of meters in size or smaller must have occurred unless all known ureilites originated from within tens of meters of the disrupted surface (an unlikely hypothesis).

Figure 5.

 Estimated mean cooling rates of ureilites showing theoretical sizes of spherical bodies with equivalent mean cooling rates at relevant thermal diffusivities. Sizes of dislodged hot “primary fragments” of UPB mantle implied by cooling rates determined from Almahata Sitta and other ureilites are much larger than asteroid 2008 TC3 or estimated size of fragments therein. This implies that significant subsequent fragmentation has occurred, and that this process was restricted to cool temperatures. The magnitude of this cold fragmentation is indicated by arrows on the figure. Our cooling rate estimates based on Fe-Mg interdiffusion in reverse-zoned olivine and pigeonite from Almahata Sitta are shown in black. An estimate of cooling rate based on exsolution wavelengths in pigeonite from Almahata Sitta by Mikouchi et al. (2010) appears as a shaded field. Cooling rates from other ureilites based on diffusion in reverse-zoned olivines by [1] Chikami et al. (1996), [2] Goodrich et al. (2001), and [3] Toyoda et al. (1986) are indicated by brackets, as is the estimate of [4] Mori and Takeda (1983) based on exsolution wavelengths in pigeonite.

Catastrophic breakup and reassembly scenarios have also been proposed to explain complex textures in mesosiderites (Haack et al. 1996a; Scott et al. 1996, 2001) and IAB iron/winonaites (Benedix et al. 1998, 2000, 2005). The UPB provides a third case study of global fragmentation of a hot differentiated (nonchondritic) asteroid. Textures in both mesosiderites and IAB iron/winonaites indicate mixing of molten metal, presumably core material, with silicate fragments. Although metal-silicate segregation did occur on the UPB (Rankenburg et al. 2008), metal-rich meteorites related to ureilites are unknown so there is no evidence that disruption of the UPB excavated as deeply as the core if one were present. Major hit-and-run collisions have been proposed as a mechanism for the disruption of all three of these parent bodies. In ureilites, we see remnants of collisional shock in the presence of diamonds and mosaicized textures (Nakamuta and Aoki 2000; Rubin 2006). In addition, the UPB possessed a possible internal mechanism to amplify disruption by the production of gaseous reduction products. This so-called “runaway smelting” could potentially result from exothermic reduction reactions triggered by a sudden drop in pressure following a major impact (Singletary and Grove 2003; Warren and Huber 2006). It is perhaps conceivable that runaway smelting would not even require an external trigger to proceed, but the abundance of shock features in ureilites and their preservation despite high equilibration temperatures implies a causal relationship between shock and the onset of rapid cooling. In the case of both mesosiderites and IAB iron/winonaites, primary fragments inferred to have been produced by disruption are observable at the specimen scale as centimeter, millimeter, and submillimeter clasts. Although the vast majority of ureilites are unbrecciated at the specimen scale, cooling rates of ureilites are several orders of magnitude faster than those proposed for either mesosiderites (104–105 °C Ma−1 at high T, but <0.5 °C Ma−1 at low T) (Ganguly et al. 1994; Ruzicka et al. 1994; Scott et al. 2001) or IAB iron/winonaites (25–200 °C Ma−1) (Herpfer et al. 1994). Cooling rates in these nonureilite examples are radically slower than can be explained by thermal diffusion rates within small fragments themselves. Heat retention by opaque dust in a debris cloud (Haack et al. 1996b; Benedix et al. 2000) and by lithic overburden after reaccretion (Benedix et al. 2000; Scott et al. 2001) have been called upon to explain this disparity. By contrast, cooling rates from ureilites seem to indicate that heat retention by opaque dust in the debris cloud had a minimal effect, and that reaccretion was delayed. Their anomalously high cooling rates give a markedly different impression of the debris cloud environment surrounding an asteroid in the aftermath of catastrophic disruption.

Post-UPB: Cold Accretion, Fragmentation, and Addition of Chondritic Material

After rapid cooling, the post-UPB thermal history is inferred to have been entirely cold. Certainly temperatures did not exceed blocking temperatures of silicate minerals for the length of time sufficient to erase observed small scale disequilibria such as reverse-zoned olivine rims. Reassembly of UPB fragments into one or more daughter asteroids must have followed sufficient lag time for cooling. At our cooling rate estimates, time spans of 1 week to 1 yr would be necessary just to cool materials below 800 °C. Love and Ahrens (1996) estimated that the majority of a catastrophically disrupted asteroid should reaccrete at shorter time scales, within hours to days. Speculatively, some combination of an especially large impact, explosive runaway smelting (Warren and Huber 2006), and the emission of CO gas from dislodged UPB fragments could have resulted in anomalous dispersion of the debris cloud. Regardless of cause, sluggish reaccretion of UPB fragments combined with a lack of subsequent heating increases the plausibility that multiple, smaller daughter asteroids may have formed in the wake of UPB disruption. This is significant because some daughter asteroids might, under certain scenarios, contain a biased sampling of the UPB. Such bias would be consistent with, but not a unique explanation for, the near exclusivity of depleted mantle materials even in polymict ureilites.

Once formed, daughter asteroids must have had little internal radiogenic heat production or must have remained small enough to minimize the thermal blanketing effect of lithic overburden. Just as in the debris cloud environment, the retention of heat by opaque dust particles surrounding UPB offspring seems to have had minimal effect. Impact heat generation could not have been a significant factor in the post-UPB history of these samples either, in agreement with the model-derived conclusions of Keil et al. (1997). The diversity in oxygen-isotopic composition of Almahata Sitta meteorites (Rumble et al. 2010) suggests that material derived from different locations on the UPB, separated by distances greater than the limits of oxygen-isotope equilibration at temperatures of partial melting, were mixed and coaccreted. It is impossible to determine for certain the mean clast size of the rubble that comprised asteroid 2008 TC3, but considering the small size of the asteroid (Scheirich et al. 2010), diversity of recovered stones (Shaddad et al. 2010; Zolensky et al. 2010), and the substantial mass lost between atmospheric entry and ground recovery (Shaddad et al. 2010), rubble within the asteroid must have been tens of centimeters and smaller. As this size range is two orders of magnitude smaller than our estimated size of fragments initially dislodged from hot UPB mantle, we infer that after breakup of the UPB subsequent fragmentation was required to produce the fragment size distribution in asteroid 2008 TC3. The timing of this latter fragmentation is unclear with respect to reassembly.

A faint chondritic chemical component has long been recognized in ureilites (Wasson et al. 1976), and chondritic clasts are identified as a minor component of several polymict ureilites (Jaques and Fitzgerald 1982; Brearley and Prinz 1992; Goodrich et al. 2004; Downes et al. 2008). Several examples of chondritic lithologies have also been recovered from the Almahata Sitta strewn field. Given the context of their location and young terrestrial ages (Shaddad et al. 2010), these chondritic fragments are interpreted to have been part of the predominately ureilitic asteroid 2008 TC3. Shaddad et al. (2010) estimated approximately 20–30% of the recovered mass of 2008 TC3 to have been comprised these chondritic materials, far in excess of the proportion of xenogenic components in polymict ureilites. The diversity of chondrite types in Almahata Sitta (Shaddad et al. 2010; Zolensky et al. 2010), as well as in regolithic polymict ureilites (Jaques and Fitzgerald 1982; Goodrich et al. 2004; Downes et al. 2008), suggests that these materials were added over a span of time permitting encounters with different asteroid or meteoroid types. Presumably accreting daughter asteroids would have been more likely to encounter predominantly ureilitic material in earlier stages and would begin to encounter more types of material as time went on. It seems conceivable that additional cold fragmentation experienced by primary UPB fragments could have resulted from bombardment by chondritic impactors on ureilitic daughter asteroids, but the relative roles of pre- and postreaccretion cold fragmentation have yet to be systematically evaluated.


Ureilitic fragments of the Almahata Sitta fall are texturally and mineralogically similar to other ureilites. Accordingly, the thermal history that they record is also similar and supportive of a disrupted UPB hypothesis, indicating that these newly recovered meteorites share a common history with other known ureilites. Work on these ureilites and others has demonstrated that reduction reactions associated with reverse-zoned rims on olivines and pyroxenes were contemporaneous with rapid cooling and, necessarily, with massive volume expansion. We assert that the sum of observations from Almahata Sitta and other ureilites indicates rapid disassembly of much or all of the UPB mantle into fragments tens-of-meters in size, rather than localized unroofing. Shock features such as microdiamonds (Steele et al. 2009; Ross et al. Forthcoming) and mosaicism (Zolensky et al. 2010) observed in Almahata Sitta combined with a potential for “runaway smelting” (Warren and Huber 2006) suggest an energetic environment. Evidence for anomalously high cooling rates preserved in mafic silicate mineral rims indicates inefficient heat retention in the debris cloud environment in the days and months following catastrophic disruption. Preservation of disequilibrium textures developed upon disruption indicates that fragments of the UPB had sufficient time to cool prior to reassembly and were never again significantly heated. The addition of a substantial and varied chondritic component to asteroid 2008 TC3 (Shaddad et al. 2010), a modern progeny of the ancestral UPB, might be the driving mechanism for decreased clast size of ureilitic components in the absence of substantial heating.

The recovery of the Almahata Sitta meteorites has also provided insight into the nature of modern ureilitic asteroids. Disintegration upon atmospheric entry suggests a poorly consolidated body, perhaps porous or with preexisting internal planes of weakness that experienced rapid ablation or mechanical failure (Jenniskens et al. 2009; Shaddad et al. 2010). Asteroid 2008 TC3 comprised many smaller fragments and the combined assemblage of recovered stones is both polylithologic and polygenetic, yet most individual specimens would appear monomict or unbrecciated without the context of other samples. The specimens of Almahata Sitta that we examined lack features of true regolithic polymict ureilites like EET 83309 and EET 87720, which are finely brecciated and contain rounded clasts, Fe,Si-metals, and xenogenic materials mixed at millimeter to centimeter scales (Mason 1986; Prinz et al. 1987; Goodrich et al. 2004) as well as solar wind implanted noble gases (Rai et al. 2003). The Almahata Sitta fall has opened the door to the possibility that the scale of brecciation and the polylithologic nature of asteroid 2008 TC3 might be representative of much of the bulk volume of modern ureilitic asteroids and not limited to their regoliths. It has also invited speculation that polylithologic and polygenetic objects might be a more common vehicle for delivery of ureilites, and perhaps other meteorite types, than previously supposed.


Acknowledgments— Thank you to the University of Khartoum, Department of Physics and Astronomy and all of the students from the University of Khartoum who aided in the recovery of Almahata Sitta, especially D. Aldhawi and R. Alderdeeri, as well as the various observers who tracked asteroid 2008 TC3. The ANSMET program and the U.S. Antarctic Meteorite Program provided recovery and allocation of other ureilites examined as part of this study. P. Abell, J. Jones, T. Mikouchi, J. Park, and P. Warren provided helpful discussion. H. Downes, S. Singletary, E. Scott, and T. Jull provided constructive comments and reviews. This work was supported in part by a grant from the NASA Cosmochemistry Program to D. W. M. M. E. Z. was supported by the Hayabusa mission. P. J. was supported by the NASA Planetary Astronomy program. A. J. R. was supported by a NERC CASE Studentship.

Editorial Handling— Dr. Edward Scott