A noble gas and cosmogenic radionuclide analysis of two ordinary chondrites from Almahata Sitta


Corresponding author. E-mail: matthias.meier@geol.lu.se


Abstract– We present the results of a noble gas (He, Ne, Ar) and cosmogenic radionuclide (10Be, 26Al, 36Cl) analysis of two chondritic fragments (#A100, L4 and #25, H5) found in the Almahata Sitta strewn field in Sudan. We confirm their earlier attribution to the same fall as the ureilites dominating the strewn field, based on the following findings: (1) both chondrite samples indicate a preatmospheric radius of approximately 300 g cm−2, consistent with the preatmospheric size of asteroid 2008 TC3 that produced the Almahata Sitta strewn field; (2) both have, within error, a 21Ne/26Al-based cosmic ray exposure age of approximately 20 Ma, identical to the reported ages of Almahata Sitta ureilites; (3) both exhibit hints of ureilitic Ar in the trapped component. We discuss a possible earlier irradiation phase for the two fragments of approximately 10–20 Ma, visible only in cosmogenic 38Ar. We also discuss the approximately 3.8 Ga (4He) and approximately 4.6 Ga (40Ar) gas retention ages, measured in both chondritic fragments. These imply that the two chondrite fragments were incorporated into the ureilite host early in solar system evolution, and that the parent asteroid from which 2008 TC3 is derived has not experienced a large break-up event in the last 3.8 Ga.


The numerous meteorites found in the Almahata Sitta strewn-field are the first ones for which the small asteroid delivering them to Earth was observed in space. This asteroid, named 2008 TC3, was detected by the Catalina Sky Survey on October 6th, 2008 to be on a collision course with Earth, and was determined to most resemble the spectroscopic class F of the Tholen taxonomy (Tholen and Barucci 1989; Jenniskens et al. 2009, 2010). The light-curve allowed Scheirich et al. (2010) to reconstruct the shape and rotation period. The asteroid entered the atmosphere 20 h after detection and broke up 37 km above ground in northern Sudan. A large, 30 × 7 km strewn-field was discovered near railway station 6 (Arabic: Almahata Sitta) by a search expedition from the University of Khartoum (Jenniskens et al. 2009; Shaddad et al. 2010), constituting, so to speak, the first “sample return mission” from a previously observed near-earth asteroid. More than 600 fragments with masses up to approximately 400 g and a total of approximately 10.7 kg have now been recovered (Shaddad et al. 2010). The majority of the meteorites have been classified as polymict anomalous ureilites (Jenniskens et al. 2010). About 20–30% of the recovered fragments are chondrites of several different types. The majority are enstatite chondrites (EL, EH); some ordinary chondrites (L, H) and one “unique” chondrite type (Horstmann et al. 2010) have also been recognized (Bischoff et al. 2010; Zolensky et al. 2010). Several arguments have been put forward to relate these chondrites to the fall of 2008 TC3. Short-lived radionuclides have been found in several chondrite fragments (including 46Sc, 54Mn, and 57Co with half-lives <1 yr; Bischoff et al. 2010). The chondrite fragments are fresh or only minimally affected by terrestrial alteration (Bischoff et al. 2010; Shaddad et al. 2010; Zolensky et al. 2010). A few of the ureilite fragments show multiple lithologies, indicating that they are part of a polymict breccia (Bischoff et al. 2010; Zolensky et al. 2010). Enstatite chondrites, which are relatively rare (making up only 1.6% of all observed falls; Grady 2000), show a high abundance among Almahata Sitta nonureilites, and would require at least one recent EH- and one recent EL-chondrite fall within the strewn-field area, which is very unlikely (Bischoff et al. 2010). The size and the position of individual nonureilite fragments within the strewn-field are consistent with those of ureilites (Shaddad et al. 2010). Sabbah et al. (2010) reported the detection of polycyclic aromatic hydrocarbons (PAHs) of ureilitic origin in some of the nonureilite fragments. In this work, we determine the preatmospheric (host) meteoroid size and cosmic ray exposure (CRE) ages of two ordinary chondrite (L4, H5) fragments. If they are indeed related to asteroid 2008 TC3, we expect the preatmospheric size and exposure age to be identical to the values reported for the Almahata Sitta ureilites (Welten et al. 2010).

Polymict ureilites with multiple xenolitihic lithologies (ordinary chondrites, R chondrites, angrites, and carbonaceous chondrites; see Downes et al. 2008; Bischoff et al. 2010; and references therein for a discussion) have been identified before, but the specific mixture of abundant enstatite chondrites next to fewer ordinary chondrite clasts is so far unique to Almahata Sitta. Similar “xenoliths” are known from a variety of other meteorite types (Bischoff et al. 2002), and a few of them have been analyzed for their noble gas content to compare their CRE ages with the matrix in which they were found embedded. Several of these xenoliths show evidence of an additional earlier irradiation phase, prior to the one in which they were delivered to Earth (Schultz et al. 1972; Schultz and Signer 1977; Bunch et al. 1979; Lorin and Pellas 1979; Pedroni et al. 1988; Wieler et al. 1989). All these xenoliths are embedded in solar-gas-rich regolith breccias, derived from asteroidal regoliths close to the surface of an asteroid. A 2π irradiation (exposure to galactic cosmic rays [GCR] within the topmost approximately 1 m of the regolith) is the most likely explanation for any inferred preirradiation phase (Wieler et al. 1989). The xenoliths, or nonureilite fragments of Almahata Sitta, are special in that the absence of solar wind-derived noble gases in any fragment of the Almahata Sitta ureilites analyzed so far (Murty et al. 2010; Ott et al. 2010; Welten et al. 2010) suggests that Almahata Sitta was never part of the regolith of the ureilite parent body, but is probably derived from an interior part of a rubble pile asteroid. Our second goal thus is to investigate a possible preirradiation of the two Almahata Sitta chondrites, to determine if clasts in such fragmental or solar-gas-free breccias also show evidence of pre-exposure to cosmic rays.

The chondrite fragments of Almahata Sitta might also give us new insight into the history and evolution of the ureilites and the ureilite parent body (UPB). Ureilites have very low concentrations of U, Th, K (and thus of the long-lived radionuclides 235,238U, 232Th, 40K), and high concentrations of primordial noble gases. The measurement of a radiogenic 4He gas retention age is not possible, and the determination of a K-Ar age is challenging. Bogard and Garrison (1994) measured 40A/39Ar-ages in four different ureilites (Novo Urei, Kenna, PCA86502, and Haverö) and reported somewhat ambiguous ages of 3.3–4.1 Ga. These are in rough agreement with an approximately 3.8 Ga Sm-Nd age measured in the light rare earth element (LREE) component of four ureilites by Goodrich and Lugmair (1995). They are also roughly coincident with the proposed late heavy bombardment (LHB). The 26Al and 53Mn chronometers, on the other hand, indicate a very early formation of the UPB (Goodrich et al. 2010). Other data point to a disruption at high temperature (1200–1300 °C; Goodrich 2004). But since the UPB would have had to be an unlikely large size (approximately lunar-sized) to remain at this temperature until the LHB (Goodrich 2004), it seems more likely that after an initial break-up, a (partially?) reaccreted UPB experienced a second break-up, or at least strong impact-related heating event, during the LHB.

The chondritic fragments found in Almahata Sitta and other polymict ureilites, however, have much higher concentrations of U, Th, and K. It seems possible that these fragments have shared at least a large part of the thermal history of the second- or third-generation ureilite parent asteroids (see Herrin et al. [2010] for a recent review of the ureilite history). If, for example, the ureilites were derived from the Hoffmeister family of asteroids, which experienced a catastrophic disruption 300 ± 200 Ma (Nesvorný et al. 2005), as suggested as a possibility based on the orbital elements of 2008 TC3 by Jenniskens et al. (2010), we expect to see this reflected in a re-setting of the gas retention age clocks at that time. A similar re-setting has been observed for many L chondrites (e.g., Bogard 1995), which experienced a catastrophic breakup event approximately 470 Ma (Korochantseva et al. 2007) that similarly led to the formation of the Gefion family of asteroids (Nesvorný et al. 2009). Accordingly, a third important goal of the present work is the analysis of the radiogenic 4He and 40Ar content of the two chondrite fragments.

Samples, Methods, and Results

Sample Selection

We received two pieces of the Almahata Sitta fragments #A100 (2.36 g) and #25 (1.68 g). Fragment #A100 is an L4 chondrite with 11.4 g original mass that was found on December 11, 2009, on a follow-up expedition toward the low-mass end of the Almahata Sitta strewn field (Shaddad et al. 2010). Fragment #25 is an H5 chondrite that was found on December 30, 2008, on the second Almahata Sitta search expedition, near the high-mass end of the strewn field. With a mass of 222 g, it is one of the very few recovered Almahata Sitta fragments with masses >200 g (Shaddad et al. 2010). Both fragments were about 70% covered by fusion crust and very fresh (unweathered) when found. No other lithologies were found in these fragments (Zolensky et al. 2010). For both pieces we received, a 0.25 g sample was separated for noble gas analysis, and the remaining mass was used for radionuclide analysis.

Noble Gas Analysis and Component Deconvolution

Samples #A100 and #25 each were divided, and each split was wrapped in aluminum foil and preheated in vacuum at 130 °C for approximately 24 h to remove adsorbed atmospheric gases. He, Ne, and Ar were extracted by pyrolysis in a Mo crucible during a single extraction at a temperature of 1800 °C for approximately 30 min. A cryo-trap cooled with liquid nitrogen was used to separate Ar from He and Ne, and to suppress interferences. Contributions from interfering species were always smaller than the ion counting error of the affected noble gas species. Blank corrections derived from these runs were typically approximately 1–2% of measured He, Ne, and Ar. A more detailed description of the noble gas analysis can be found in Wieler et al. (1989).

The noble gas results are shown in Table 1. For the deconvolution of the trapped and cosmogenic noble gas components, we used the values for trapped and cosmogenic components given by Ott (2002) and Wieler (2002), respectively. The measured 20Ne/22Ne ratio of approximately 0.84–0.86 implies an almost entirely cosmogenic composition for Ne in both samples. From this, we conclude that 3He is almost entirely cosmogenic in both samples. Using (4He/3He)cos = ∼6 (Alexeev 1998), we estimate that approximately 11% of the total 4He is cosmogenic, the rest is assumed to be radiogenic. The 36Ar/38Ar ratios of approximately 4.2 (#A100) and approximately 3.1 (#25) indicate a considerable amount of trapped Ar. For the deconvolution of Ar, we assume that the cosmogenic and trapped end-members have a 36Ar/38Ar of 0.65 and 5.3, respectively. Corrections for trapped 38Ar are 49–54% for #25 and 76–77% for #A100. Contributions from the decay of neutron-capture-produced 36Cl (36Clnc) to 36Ar, as discussed later, are on the order of approximately 1% of total 36Ar. These were subtracted before deconvolution. Only a minor amount (<0.1%) of 40Ar can be of trapped/cosmogenic origin, and as both fragments are very fresh, we assume that they do not contain any trapped atmospheric Ar. We conclude that 40Ar is entirely radiogenic.

Table 1.  . Noble gas results of Almahata Sitta chondrites.
  1. Ion counting errors, uncertainty of the blanks and of sample mass are about 0.1%. The uncertainty of mass spectrometer sensitivity is <1% for ratios and <5% for concentrations.

TypeL4 chondriteH5 chondrite
Mass (mg)162.365.84183.267.80
Measured concentrations (10−8 cm3 STP g−1)
Measured isotopic ratios
Cosmogenic concentrations (10−8 cm3 STP g−1)
Radiogenic concentrations (10−8 cm3 STP g−1)
 40Arrad = 40Armeas8094716774807593

Cosmogenic Radionuclides and Chemical Analysis

For the analysis of the cosmogenic radionuclides 10Be (half-life = 1.36 × 106a), 26Al (7.05 × 105a), and 36Cl (3.01 × 105a), we separated the metal and stone fractions from both chondrites. The purified metal fraction was dissolved along with Be, Al, and Cl carrier in HNO3, while the stone fraction was dissolved, in the presence of Be and Cl carrier, in HF/HNO3. After dissolution, a small aliquot was taken for chemical analysis of major elements by atomic absorption spectrometry (Perkin Elmer AA3300), and of Th by ICP-MS (Thermo Electron X Series II, following procedures described in Friedrich et al. 2003). Beryllium, Al, and Cl were separated using procedures previously described in Welten et al. (2010). AMS measurements of 10Be, 26Al, and 36Cl were performed at PRIME Lab, following procedures described previously by Sharma et al. (2000). After subtracting blank levels, we normalized the measured isotopic ratios to 10Be, 26Al, and 36Cl AMS standards (Sharma et al. 1990; Nishiizumi 2004; Nishiizumi et al. 2007). The cosmogenic radionuclide results are given in Table 2. The quoted errors represent only 1σ uncertainties in the AMS measurements. As the measured U concentration was very close to the background and the measured Th concentration is within error compatible with chondritic values, we use the chondritic U and Th concentrations given by Wasson and Kallemeyn (1988) for the calculation of a U, Th-He age.

Table 2.  . Major element composition and cosmogenic radionuclide activities.
  #A100-metal#A100-stone#A100 bulk#25-metal#25-stone#25 bulk
Mass (mg)44.3123.577.2133.0
Major element composition, in wt%
Cosmogenic radionuclide activities, in dpm kg−1
 10Be3.99 ± 0.1021.8 ± 0.7020.5 ± 0.62.92 ± 0.1020.2 ± 0.6017.1 ± 0.5
 26Al3.38 ± 0.2068.7 ± 2.663.7 ± 2.42.37 ± 0.1266.7 ± 2.955.0 ± 2.4
 36Cl19.4 ± 0.510.5 ± 0.211.2 ± 0.214.4 ± 0.48.1 ± 0.29.2 ± 0.2
 36Clnc 3.9 ± 0.9 2.3 ± 0.8


Preatmospheric Size and Shielding Depth

If the two chondrite fragments were delivered to Earth together with the Almahata Sitta ureilites, the preatmospheric radius derived from each of these fragments must be identical to the one derived from the Almahata Sitta ureilites. The preatmospheric radius of the meteoroid that delivered the Almahata Sitta ureilites (asteroid 2008 TC3) is estimated to be approximately 300 g cm−2, equivalent to a mass of 20,000–50,000 kg (Welten et al. 2010). One could argue though that two (or more) large strewn fields happen to overlap. However, fewer than approximately 10 meteoroids with preatmospheric masses >20,000 kg enter the Earth’s atmosphere every year (Bland and Artemevia 2006), and both ureilites and nonureilites in the Almahata Sitta strewn-field are very fresh. Therefore, an accidental association of fragments from two (or more) strewn fields can essentially be excluded.

There are no cosmic ray irradiation models tailored to chondritic clasts embedded in a ureilitic host. Welten et al. (2010) interpreted the noble gas concentrations and cosmogenic radionuclide activities of Almahata Sitta ureilites using the Leya and Masarik (2009) model developed for carbonaceous chondrites, assuming a density of 2.25 g cm−3. Due to the high carbon content (approximately 2 wt%; Steele, unpublished data) and the low density (Kohout et al. 2010; Shaddad et al. 2010) of the Almahata Sitta ureilites, this model may yield a better approximation to the flux density and energy distribution of secondary particles in ureilites than the ordinary chondrite model. Here, we use the noble gas and cosmogenic radionuclide production rates from each target element as calculated in the carbonaceous chondrite model by Leya and Masarik (2009). To calculate the total production rate for each cosmogenic species, we adopt the measured bulk concentration for the two fragments in Mg, Al, Ca, Mn, Fe, and Ni, and average L, H chondrite composition for C, O, Na, and Si (Wasson and Kallemeyn 1988), thereby simulating chondritic samples embedded in a carbonaceous chondrite matrix.

The activity of 10Be in the stone fraction relative to the 10Be activity in the metal fraction (e.g., Welten et al. 2001, 2003) is an excellent indicator of preatmospheric size. Figure 1a shows a plot of the 10Be activity in the stone fraction versus the 10Be activity in the metal fraction of the two chondrite fragments. The 10Be activities yield well-defined, usually nonintersecting curves for any meteoroid size (see Fig. 1a). To derive these curves, we use the LCS model by Masarik and Reedy (1994b), which has proven to be useful for large objects (Welten et al. 2003, 2011). The 10Be activities in the stone and metal fractions of the two chondrite fragments plot within the modeled range for an approximately 300–400 g cm−2 object. This radius is indistinguishable to the approximately 300 g cm−2 preatmospheric radius determined for the Almahata Sitta ureilite fragments by Welten et al. (2010). The concentrations of cosmogenic 10Be in the metal phase of #25 are approximately 30% lower than in #A100 (see Table 2), indicating higher shielding for fragment #25. On the basis of 10Be depth profiles for an object with a radius of approximately 300 g cm−2, we derive shielding depths of 30–50 g cm−2 for #A100 and 80–150 g cm−2 for #25 (Figs. 1b and 2c). The higher shielding of fragment #25 is also confirmed by a lower 22Ne/21Ne ratio (1.08 versus 1.10 for #A100) and a lower 3He/21Ne ratio (3.55 versus 3.85 for #A100), which both decrease monotonically with increasing shielding depth within the relevant shielding range (30–150 g cm−2).

Figure 1.

 Beryllium-10 in stone versus metal fraction and resulting shielding depths. (a) Beryllium-10 concentrations in dpm kg−1 measured in the stone versus metal fraction of the two Almahata Sitta chondrites (#25, H5, and #A100, L4), plotted together with the modeled concentrations based on the LCS model for chondrites by Masarik and Reedy (1994a) for three different meteoroid sizes (solid lines: 440 g cm−2, long-dashed lines: 370 g cm−2, short-dashed lines: 290 g cm−2) of different chemistry (gray: H; black: L). The meteoroid size for the Almahata Sitta ureilites is approximately 300 g cm−2 (Welten et al. 2010). Errors are 1σ. Beryllium-10 in the metal phase (in dpm kg−1) versus shielding depth is shown in panel (b) for L, and panel (c) H chondrite chemistry. From the concentrations of 10Be in the two chondrite samples, shielding depths of 30–50 g cm−2 and 100–150 g cm−2 were derived for #A100 (b) and #25 (c), respectively.

Figure 2.

 Concentrations of cosmogenic He, Ne for Almahata Sitta samples. The concentration of cosmogenic 3He (white), 21Ne (light gray), and 38Ar (dark gray) from the four subsamples of the two chondrite fragments (top), and from four ureilite fragments of Almahata Sitta (from Welten et al. 2010), and the cosmic ray exposure ages (bottom) for 21Ne (light gray) and 38Ar (dark gray). Note that the concentrations for 3He and 38Ar have been scaled with a factor of 0.3 and 8, respectively, to make the depiction more readable. No cosmogenic 38Ar concentration for the ureilite samples was given in Welten et al. (2010); see Shaddad et al. (2010) for more information on the sample/fragment numbers.

From 26Al-data, a somewhat smaller preatmospheric size is derived, especially for #A100. We do not, however, consider the 26Al results a reliable indicator of preatmospheric size in this instance. The 26Al/10Be ratios of 0.85 ± 0.05 and 0.81 ± 0.05 in the metal phase of #A100 and #25 are anomalously high. Typical ratios are 0.70 ± 0.03 in irons and metal phases of chondrites and stony irons (Aylmer et al. 1988; Lavielle et al. 1999; Albrecht et al. 2000). The 36Cl/10Be ratio in the metal phase and the 26Al/10Be ratio in the stone phase of both chondritic fragments are in the typical range for large chondrites (e.g., Welten et al. 2003). It seems likely that the high 26Al-concentration in the metal phase is not related to the shielding conditions or a short CRE age, but rather to high P content (0.1–0.2 wt%) in the metal phase (a high S content can be excluded on the basis that troilite was leached in 0.2N HCl during sample preparation).

We estimate the amount of 36Clnc by subtracting the activity of 36Cl expected from spallation (36Clsp) in the stone phase from the measured 36Cl activity in the stone phase. Chlorine-36 from spallation is calculated as 36Clsp = 36Cl(metal) × [1.1 × Fe + a × Ca + b × K]. Here, a is the production rate ratio of 36Clsp produced from Ca relative to that from Fe, which can be estimated using the 10Be(stone)/10Be(metal) ratio (resulting in a = 13.2 for #A100, 15.9 for #25), and b = 1.8 × a, based on Welten et al. (2001). This results in 36Clnc activities of 3.9 and 2.3 dpm kg−1 for #A100 and #25, respectively. Together with the shielding depths as calculated above, 36Clnc can then be used to derive the Cl concentrations in the bulk samples, based on the model by Spergel et al. (1986). This results in a Cl concentration of 60–80 ppm for #A100, and 10–20 ppm for the H chondrite fragment #25. These are both within the range observed by Garrison et al. (2000). As neutron-capture reactions are only important in objects with a large preatmospheric radius, the detection of 36Clnc is a further confirmation that the precursor of the two chondrite fragments was a large object. Due to the relatively low amounts of 36Clnc detected, corrections of the measured 36Ar concentrations for contributions of 36Ar from the decay of 36Clnc are nearly negligible.

Cosmic Ray Exposure Ages

As the cosmogenic radionuclides 10Be, 26Al, and 36Cl show activities near equilibrium values for moderately shielded ordinary chondrites, they can only provide a lower limit (>5 Ma) of the CRE age. The CRE ages can be determined from the noble gas isotopes 3He, 21Ne, and 38Ar. Each of these isotopes has contributions from noncosmogenic and cosmogenic (3Hecos, 21Necos, 38Arcos) sources. The 3Hecos and 21Necos concentrations in both chondritic fragments are similar or slightly higher than their counterparts in fragments from the ureilitic host (20–30 × 10−8 cm3 STP g−1 and 6–8 × 10−8 cm3 STP g−1 for 3Hecos and 21Necos, respectively, see Fig. 2) (Welten et al. 2010). This similarity in concentrations suggests that the two chondrite fragments have been exposed for similar time spans as the ureilites, approximately 20 Ma (Welten et al. 2010).

To determine the precise 21Ne exposure ages of the two chondrite fragments, we first need to determine the 21Ne production rate. The method developed by Eugster (1988) is not applicable for meteoroids of large (>50 cm radius) preatmospheric size (Leya and Masarik 2009). Instead, we use the 21Ne/26Al-production rate ratio, P(21)/P(26), which is almost insensitive to the size of the meteoroid and approaches a stable value for shielding depths >100 g cm−2 (e.g., Graf et al. 1990). The production rate ratio P(21)/P(26) for meteorites with approximately 300 g cm−2 radius increases from approximately 0.0052 (in units of 10−8 cm3 STP g−1 Ma−1 dpm kg−1) near the surface to 0.0057 for depths >100 g cm−2 (Fig. 3) for H chondrite chemistry. We adopt P(21)/P(26) ratios of 0.0054 for #A100 and 0.0057 for #25 (an error of 10% is added and propagated on all production rates), respectively. Taken together with the measured 26Al-concentrations (Table 2), we obtain 21Ne-production rates of 0.34 × 10−8 cm3 STP g−1 Ma−1 and 0.31 × 10−8 cm3 STP g−1 Ma−1. This results in CRE ages of 22 ± 4 and 21 ± 3 Ma for the two chondrite fragments #A100 and #25, respectively (Fig. 2, Table 3). These CRE ages are, within error, indistinguishable from the average CRE age of 19.5 ± 2.5 Ma for the Almahata Sitta ureilites that was determined using the same 21Ne/26Al method (Welten et al. 2010). We conclude that the two chondrite fragments were indeed delivered to Earth within the same parent object as the Almahata Sitta ureilites, i.e., asteroid 2008 TC3.

Figure 3.

21Ne/26Al production rate ratio. Calculated production rate ratio of 21Ne/26Al in an H-chondrite (fragment #25) embedded in a carbonaceous-chondrite like matrix and, based on the irradiation model by Leya and Masarik (2009), for a meteoroid of 345 g cm−2 total shielding. Note that the production rate ratio is nearly insensitive to shielding (approximately 10% change over 200 g cm−2 shielding).

Table 3.  . Cosmic ray exposure and gas retention ages (in Ma).
FragmentT21/26T38T38pre-expR4 (chond)R40 (meas)
#A100, average22 ± 440 ± 4>16 ± 337704380
#25, average21 ± 331 ± 3>8 ± 237604710

Cosmogenic 38Ar: Evidence for Preirradiation?

While the determination of 38Arcos in the ureilite fragments was not possible due to large contributions from trapped Ar and large variations in Ca content (Welten et al. 2010), this problem does not affect the two chondrite fragments. We report 38Arcos concentrations of approximately 1–1.2 × 10−8 cm3 STP g−1 for both fragments in Table 1 and Fig. 2. The irradiation model by Leya and Masarik (2009) yields 38Ar production rates of 3.1 × 10−10 cm3 STP g−1 Ma−1 for #A100 and 3.7 × 10−10 cm3 STP g−1 Ma−1 for #25, resulting in nominal 38Ar CRE ages of 40 ± 4 and 31 ± 3 Ma, respectively. The 38Ar ages of the two chondrite fragments are significantly larger than their 21Ne/26Al-ages and the 19.5 ± 2.5 Ma age of the Almahata Sitta ureilites.

If the measured 38Arcos concentration had been acquired in only 19.5 ± 2.5 Ma, it would require 38Ar production rates significantly higher than any 38Ar production rate (for any shielding and meteoroid size) predicted by the model by Leya and Masarik (2009). High production rates require Ca concentrations on the order of approximately 2.6% (versus 0.98% measured) for #A100 and approximately 2.1% (versus 1.16% measured) for #25. It is known that approximately 100 mg sized chondrite samples can have quite variable Ca concentrations. However, to find such high values simultaneously in both splits of both fragments seems very unlikely. Adopting average chondritic Ca values (from Wasson and Kallemeyn 1988) lowers the nominal CRE age only to 34 ± 3 and 28 ± 2 Ma for the #A100 and #25 fragments, respectively. This age is still longer than the 21Ne/26Al-age of the fragments or the Almahata Sitta ureilites. A loss of 30–50% of the 21Necos during the latest irradiation phase (as part of 2008 TC3) can also be excluded as 3Hecos/21Necos, which would probably be fractionated more than 21Necos/38Arcos, is within 15% of the typical value expected for chondrites under the given shielding conditions. Finally, not correcting 36Ar for the contribution of 36Clnc would raise the 38Arcos concentration by no more than 3%. It is possible that the high 38Arcos concentrations reflect a longer total exposure to GCR than implied by the 21Ne/26Al data. Conceivably, there was an earlier irradiation phase, followed by a loss of He and Ne (e.g., upon incorporation into the ureilitic matrix), prior to the last exposure to GCR as parts of 2008 TC3.

Preirradiation of “xenoliths” in brecciated meteorites has been observed before (e.g., Schultz et al. [1972] in Weston; Schultz and Signer [1977] in St. Mesmin; Lorin and Pellas [1979] in Djermaia; Pedroni et al. [1988] in Kapoeta; Wieler et al. [1989] in Fayetteville). In most cases, this was attributed to 2π irradiation in an asteroidal regolith, which we can exclude for Almahata Sitta as it is not a solar-gas-rich regolith breccia (Murty et al. 2010; Ott et al. 2010; Welten et al. 2010; this work). Therefore, the two chondrite fragments were probably exposed to GCR in space (4π) before their incorporation into the ureilite host. The absolute time of preirradiation in space depends on their size, but the maximum production rate as individual objects (based on the Leya and Masarik [2009] model for ordinary chondrites) for #A100 is approximately 4.1 × 10−10 cm3 STP g−1 Ma−1, and approximately 4.9 × 10−10 cmSTP g−1 Ma−1 for #25, resulting in minimum preirradiation CRE ages of 16 ± 3 and 8 ± 2 Ma for #A100 and #25, respectively. This is significantly longer than the collisional lifetime of 2–3 Ma for fragments of a few cm size (Farinella and Vokrouhlicky 1999). The two chondritic fragments are thus likely to be fragments of larger meteoroids that broke up upon impact or incorporation into the ureilite host.

Thermal History

Both chondrite fragments contain significant amounts of radiogenic 4He and 40Ar (Table 1). Assuming a chondritic concentration in U and Th (13 ppb U and 43 ppb Th for the L chondrite, 12 ppb U and 42 ppb Th for the H chondrite, Wasson and Kallemeyn 1988), the concentrations of the radiogenic noble gases yield average 4He retention ages of 3.8 Ga for both chondrite fragments, or possibly up to 0.37 Ga longer if indeed approximately 15% of the He was lost at a late stage (e.g., the breakup of 2008 TC3 in Earth’s atmosphere), as discussed before (Fig. 4). The lower figure corresponds roughly to the LHB period. The Ar-retention ages calculated from the measured bulk K concentration of 955 and 790 ppm for the #A100 and #25 fragment, respectively, are higher: 4.4 Ga for #A100 and 4.7 Ga for #25 (Fig. 4). While simple U-Th-He and K-Ar retention ages are always somewhat ambiguous, we conclude that both chondrite fragments have not experienced a complete loss of radiogenic He and/or Ar for most of the history of the solar system. Also, given the significantly lower concentration of U, Th, and K in ureilites, it is not possible that these radiogenic noble gases were somehow inherited from the ureilite host asteroid. These observations imply that at least the ureilite parent body of which they were part never experienced a severe breakup event that reset the radiogenic clocks of all or most constituent materials, e.g., the breakup event of the L chondrite parent body did for a significant fraction of the L chondrites. This also implies that asteroid 2008 TC3 is not derived from the Hoffmeister family of asteroids, which experienced a catastrophic disruption 300 ± 200 Ma (Nesvorný et al. 2005).

Figure 4.

4He, 40Ar gas retention ages. Radiogenic retention ages in Ga for 4He and 40Ar in all four subsamples. The black columns represent the U,Th-He gas-retention age resulting from a chondritic U concentration, with the upper error bar accounting for a 15% He-loss. The gray columns represent the K-Ar age.

In the previous section, we proposed that an earlier irradiation phase might have produced cosmogenic noble gases in the two chondritic fragments, of which only 38Arcos remains. The putative event that led to the loss of He, Ne cannot have been more recent than 3.8 Ga according to the U, Th-He age measured in both fragments. The preirradiation phase and thus presumably also the incorporation of the chondrite fragments into the ureilite host must have taken place ≥3.8 Ga ago, at or before the time of the LHB. As some ureilites have Ar-Ar-ages (Bogard and Garrison 1994) similar to the LHB, and/or show a Sm-Nd-isochron of approximately 3.8 Ga (Goodrich and Lugmair 1995), it seems likely that the (second-generation, or reassembled) ureilite parent body experienced a major disruption around the time of the LHB. The ureilitic fragments of this breakup event later reassembled into (third-generation) asteroids (Goodrich 2004; Herrin et al. 2010; Jenniskens et al. 2010), thereby incorporating some nonureilitic fragments (like the L and H chondrites analyzed here), which were perhaps themselves dislodged by impacts on their respective parent bodies during the LHB. These nonureilitic fragments may thus represent a snapshot of the formation environment of the asteroids, which later became the parent bodies of the ureilites delivered to Earth today (Bischoff et al. 2010; Jenniskens et al. 2010; Gayon-Markt et al. 2011).

A Trapped Ureilitic Ar Component?

Both chondrite samples contain trapped 36Ar in concentrations that are surprisingly large for equilibrated chondrites (approximately 22.6 × 10−8 cm3 STP g−1 for #A100 and 6.5 × 10−8 cm3 STP g−1 for #25). Carbon-rich separates from ureilites have high concentrations of 36Ar, of up to approximately 80,000 × 10−8 cm3 STP g−1 (Goebel et al. 1978). Bulk Almahata Sitta ureilite samples contain up to 1390 × 10−8 cm3 STP g−1 trapped 36Ar (Welten et al. 2010). We thus explore the possibility that the trapped 36Ar found in the two chondritic fragments is of ureilitic, i.e., Almahata Sitta host origin.

A significant contribution of 36Ar from the decay of neutron-capture 36Cl can be excluded, since 2–4 dpm kg−1 36Clnc only yield 1–1.7 × 10−9 cm3 STP g−1 over an approximately 20 Ma exposure age. This represents only about 1–2% of the total 36Ar (or up to double this amount when taking into account a preirradiation phase under similar shielding conditions).

An atmospheric origin of the 36Ar would, by proportional addition of atmospheric 40Ar, also require a significant reduction of the approximately 4.6 Ga K-Ar gas retention ages, making them shorter than the U,Th-He ages. The atmospheric and radiogenic 40Ar would then also fortuitously add up to a total approximately 4.6 Ga retention age, for both fragments. We thus conclude that a large atmospheric contribution is unlikely.

In the compilation of Schultz and Franke (2004), we found only one of 77 listed L4 chondrite samples (from 28 different meteorites) with measured He, Ne, and Ar concentrations that has a higher trapped 36Ar concentration than the #A100 fragment. This meteorite, however, is a solar-gas-rich regolith-breccia (20Ne/22Ne > 10, 20Ne/36Ar > 10). Likewise, only 14 of 435 listed H5 samples (from 257 different meteorites) with measured He, Ne, and Ar concentrations have higher trapped 36Ar concentrations than the average of the two splits of the #25 fragment. Six of these meteorites are solar-gas-rich regolith breccias, and the remaining meteorites all have an atmosphere-like 20Ne/36Ar > 0.2. The trapped 20Ne/36Ar ratios of approximately 0.01–0.02 for both Almahata Sitta chondrite fragments are, however, an order of magnitude lower than this.

Many ordinary chondrites contain trapped noble gases of phase Q, which has a 20Ne/36Ar ratio of approximately 0.04 (Busemann et al. 2000), within a factor of 2 of the trapped 20Ne/36Ar ratio in the two Almahata Sitta chondrite fragments. However, none of the meteorites from the compilation by Schultz and Franke (2004) contained similarly high concentrations of 36Ar of phase Q origin. A possible alternative to a phase Q origin is a “ureilitic” origin.

Goebel et al. (1978) identified a distinct “ureilitic” noble gas elemental abundance pattern in carbon-rich separates of three ureilites. This pattern is similar to the phase Q pattern observed in many carbonaceous and ordinary chondrites (e.g., Busemann et al. 2000), but has higher 36Ar/84Kr and 36Ar/132Xe ratios (Rai et al. 2003), and lower 4He/36Ar and 20Ne/36Ar ratios (Ott 2002), i.e., it is rich in 36Ar, compared with other components. A ureilitic, i.e., Almahata Sitta ureilite origin for the measured noncosmogenic 36Ar, may be a viable explanation. As we have only analyzed bulk samples in a single temperature step, we can only speculate about the carrier of this ureilitic component within the chondritic samples. In ureilites, the carbon-rich vein material is extremely rich in trapped 36Ar (Goebel et al. 1978) and it was shown by Fukunaga et al. (1987) that diamonds are the main carriers of the ureilitic component. It is conceivable that some of that material was mobilized and deposited on or within the chondrite samples over their long, >3.8 Ga, residence time in the ureilitic host matrix. The detection of polycyclic aromatic hydrocarbons (PAHs) of ureilitic origin in fragment #25 by Sabbah et al. (2010) led these authors to a similar conclusion.

In any case, if the trapped 36Ar in the chondrite fragments indeed derives from a ureilite source (irrespective of the nature of the carrier phase), this is yet another strong argument that the Almahata Sitta chondrite and ureilite fragments were once part of the same asteroid.


On the basis of an analysis of noble gases (He, Ne, Ar) and cosmogenic radionuclides (10Be, 26Al, 36Cl) of two chondrite fragments (#A100, L4 and #25, H5) found in the Almahata Sitta strewn-field, we can make the following five conclusions:

The two chondrite fragments derive from a meteoroid with a preatmospheric radius of approximately 300 g cm−2, i.e., within error identical to the size of asteroid 2008 TC3 that produced the Almahata Sitta strewn-field. Shielding depths of 30–50 and 100–150 g cm−2 are calculated for the L4 and H5 fragment, respectively. Although the concentrations of neutron-capture 36Cl are relatively low (2–4 dpm kg−1) they are consistent with a large preatmospheric size for the precursor meteoroid, thus providing an independent confirmation.

The cosmic-ray exposure ages of the two fragments, as measured with the 21Ne/26Al method, are 22 ± 4 and 21 ± 3 Ma for the L4 and H5 fragment, respectively. This is, within error, identical to the 21Ne/26Al-based cosmic ray exposure age of 19.5 ± 2.5 determined for the ureilites by Welten et al. (2010), independently confirming that the ordinary chondrite fragments were delivered to Earth within the same object as the Almahata Sitta ureilites, i.e., asteroid 2008 TC3.

For 38Arcos, a longer cosmic ray exposure age of 40 ± 4 and 31 ± 3 Ma is calculated. This can be interpreted as an additional irradiation phase (followed by loss of He and Ne) of at least 16 ± 3 and 8 ± 2 Ma for the L4 and H5 fragments, respectively, when assuming 4π irradiation in space before incorporation of the fragments into the ureilite host.

Using typical chondritic U, Th, and measured K concentrations, we derive U, Th-He ages of 3.8 Ga, and a K-Ar age of 4.4 Ga, and 4.7 Ga for the L4 and H5 fragment, respectively. These results also indicate that the ureilite parent body (bodies?), in contrast e.g., to the L chondrite parent body, did not experience a catastrophic disruption resulting in the loss of radiogenic He, Ar during the last 3.8 Ga.

The trapped Ar content of the two samples is large for equilibrated chondrites. This could be the result of the admixture of small amounts of Ar-carrying phases of ureilitic origin into (or onto) the chondritic fragments, providing another link between the Almahata Sitta chondrite and ureilite fragments.

Acknowledgments–– This work was supported by the Swiss National Science Foundation, NASA Cosmochemistry and the Planetary Astronomy program. We thank the reviewers H. Downes and I. Leya for their helpful comments, and C. Alwmark for his comments on an early version of this manuscript.

Editorial Handling— Dr. Ian Lyon