The elemental composition of Almahata Sitta


Corresponding author. E-mail:


Abstract– We quantified up to 60 elements in four individual fragments (#4, #7, #15, and #47) of the Almahata Sitta meteorite, which entered Earth’s atmosphere on October 7, 2008. Almahata Sitta is indisputably the product of fragments of asteroid 2008 TC3 surviving passage through Earth’s atmosphere. Each of the four analyzed fragments has a ureilitic composition based on comparisons with literature data. The analyzed fragments represent each of the major lithologies of Almahata Sitta. Highly refractory siderophile elements are similarly enriched in all fragments of Almahata Sitta, while volatile elements show more compositional variability. The abundances of rare earth elements (REE) only reach CI values in the case of a single Almahata Sitta fragment (#15). More typically, REE abundances in Almahata Sitta are between 0.01 and 0.3 × CI. None of the fragments displays the classic V-shaped REE pattern that are often found in ureilites. Given that Almahata Sitta was collected shortly after its fall, these REE compositional results indicate that the V-shaped REE pattern found in many ureilites may be a result of extended terrestrial residence times.


The discovery, observation of asteroid 2008 TC3’s approach toward Earth, and recovery of associated asteroidal fragments in the form of the Almahata Sitta meteorite, has provided an unprecedented connection between asteroids and meteorites. Initial petrographic, chemical, and isotopic examination of the resulting meteorite identified it as an anomalous polymict ureilite (Jenniskens et al. 2009; Rumble et al. 2010; Zolensky et al. 2010). Ureilites are enigmatic carbon-rich ultramafic achondrites with origins that reflect early high-temperature planetary differentiation processes curiously combined with a complement of characteristics that are trademarks of more primitive (chondritic) materials (for reviews see Goodrich 1992 and Mittlefehldt et al. 1998). Initial chemical analysis of Almahata Sitta was performed on a small (approximately 50 mg) sample of single fragment (#7) for trace elements only (Jenniskens et al. 2009). In this work, we expand the investigation in terms of both number and petrographic type of fragments investigated and elements analyzed. The augmented compositional results again show that the three additional fragments, along with the original fragment #7, are unquestionably ureilites when compositions are compared to literature values. However, as we will see, only occasionally does Almahata Sitta clearly exhibit the compositional trademarks typically associated with the handful of other known polymict ureilites. Since the various fragments of Almahata Sitta were collected between only 2 and 5 months after their documented fall, the elemental contents are expected to be relatively unaltered by terrestrial weathering processes, which are a concern for the majority (>97%) of ureilites, which are finds collected from cold Antarctic ice fields or hot deserts.



To date, we have studied portions of four individuals of the Almahata Sitta meteorite. The first sample consisted of 51.5 mg of fragment #7. Trace element results for this fragment were reported in Jenniskens et al. (2009). Unfortunately, we are unable to present any additional compositional information in the form of major or minor elements for fragment #7. Following up on this work, we analyzed 35.5 mg of crumbs of fragment #15 and about 200 mg of fragment #47. The latter sample was homogenized by grinding in a clean agate mortar and pestle and analyzed by us in two separate aliquots of 103.2 and 94.1 mg. In addition, we received two larger chips of fragment #4. Each chip was about one quarter of a gram in mass. Both chips were individually homogenized as above and, in total, these pieces yielded five aliquots of 126.6, 110.5, 86.4, 94.9, and 91.2 mg. Each of these aliquots was individually processed for analysis and to minimize the effects of sample inhomogeneity in our compositional reporting, we report the mean of all analytical aliquots processed for each fragment here.

The four samples were derived from meteorites with different macroscopic textures, representing just a few of many faces of Almahata Sitta. Almahata Sitta’s dominant lithologies include: a pyroxene-dominated, very porous, highly reduced lithology; a pyroxene-rich compact lithology; and an olivine-dominated compact lithology (Zolensky et al. 2010). Sample #7 was a representative of the fine grained pyroxene-rich material and possesses macroporosity resulting from partially sintered grains (Jenniskens et al. 2009; Zolensky et al. 2010). Samples #4 and #15 were representatives of the pyroxene-abundant compact lithology (Zolensky et al. 2010). Sample #4 was fragile, course grained, and homogeneously dark in appearance, with a flat reflectance spectrum between 0.35 and 2.5 μm (Hiroi et al. 2010). The pyroxene content was at 7–12% (Sandford et al. 2010). Sample #15, also course grained, but more cohesive, with approximately 7% pyroxene, was taken from a nearly completely fusion-crusted meteorite with a density of 3.11 ± 0.02 g cm−3. Although the mineralogy and petrography were not investigated in detail (Zolensky et al. 2010), analyses of sample #47 by Sandford et al. (2010) suggest that fragment #47 is representative of the olivine-dominated compact lithology (99%, versus 1% pyroxene). It contained only a weak pyroxene band at 0.9 μm and olivine absorption <0.5 μm, between 0.9 and 1.6 μm, and around 2 μm (Hiroi et al. 2010). Sample #47 had broken early into small fragments and may have been slightly weathered, with spots of brown limonite forming on metal fractions. Its density was measured at 2.96 ± 0.05 g cm−3 (Shaddad et al. 2010), also suggesting grouping in the compact lithology.

Analytical Methods

We determined the bulk composition of Almahata Sitta fragments by analyzing solutions derived from acid-dissolution of homogenized powders using quadrupole-based inductively coupled plasma–mass spectrometry (ICP-MS). Concentrations of trace elements were determined using methodology described in Friedrich et al. (2003) using a Thermo Electron X Series II ICP-MS. In addition, major and minor elements Na, Mg, Al, P, K, Ca, Cr, Fe, and Ni were determined on diluted aliquots of dissolved meteorite using methodology and instrumentation described in Wolf et al. (2005). The highly volatile trace elements Cd, In, Tl, and Bi were also determined using methodology and instrumentation described in Wolf et al. (2005). To assess the accuracy and precision of our methodology on terrestrial ultramafic rocks with elemental compositions similar to ureilites, we refer the reader to the Appendix.

Acid dissolution procedures involved a HF-HNO3 microwave digestion step followed by heating to incipient dryness at temperatures no greater than 70 °C. The subsequent addition of perchloric acid followed by heating at the same temperature as the previous step was then used to achieve nearly complete dissolution. Nevertheless, some carbonaceous residue remained in the case of our samples of fragments #4, #15, and #47. Our sample of fragment #7 did not appear to contain noticeable amounts of this perchloric-insoluble carbonaceous residue. Such carbonaceous matter is known to be present in ureilites in the form of graphite and diamond (Rai et al. 2002) and the lack of carbonaceous matter points to an inhomogeneous sampling of our sample of fragment #7 (see the Results section). The relative density of the residue was clearly lower than that of chromite, another mineral that can be resistant to repeated harsh acid digestion procedures. Chromite is a phase rarely found in the ureilites (Mittlefehldt et al. 1998).


We present compositional data for up to 60 elements in four different fragments of Almahata Sitta in Table 1. As stated in the Samples section, we received two fractions of Almahata Sitta fragment #4. Because each of these aliquots gave nearly identical results, we report the mean concentrations of the two in Table 1. When concentrations of major and minor elements are used to estimate variability due to sampling, within-fragment compositional variability ranged from 0.81% to 24% expressed as %-relative standard deviation (%-RSD). Mg and Fe varied at 0.94 and 5.8%-RSD, respectively. Al and Ca demonstrated relatively higher variabilities at 13 and 24%-RSD, respectively. Variation of minor elements Na, P, Cr, Mn, and Ni between these two chips were all <10%-RSD. Both K results were below limits of detection. The compositional variability between chips of fragment #4 is comparable to that observed in two adjacent chips of polymict ureilite Elephant Moraine (EET) 83309 (Warren and Kallemeyn 1989). Our analytical precision (reproducibility) for USGS reference materials is discussed in the Appendix.

Table 1.   Bulk composition of four fragments of the Almahata Sitta ureilite.
ElementUnitFragment #7aFragment #4Fragment #15Fragment #47
  1. NA = not analyzed.

  2. aTrace element results reproduced from Jenniskens et al. (2009) for sake of comparison. Small fragments of ureilites are relatively inhomogeneous for sample masses analyzed––see the Methods section. For an estimate of intra-fragment homogeneity in the case of fragment #4, all sample replicates are within 24% relative standard deviation except where errors are listed. Suspiciously high values (for Ag) shown in italics (see text).

Liμg g−1  1.7 1.0 0.95 1.3
Namg g−1NA 0.26 0.76 0.26
Alwt%NA 0.262 0.477 0.148
Pmg g−1NA 0.74 1.6 0.49
Kmg g−1NA<0.03<0.03<0.03
Cawt%NA 1.05 4.23 0.87
Scμg g−1  9.0 9.220.2 7.8
Tiμg g−1 274175 676 126
Vμg g−1 84 87 118  88
Crmg g−1NA 4.18 3.34 5.02
Mnmg g−1  3.06 3.72 3.66 2.70
Fewt%NA10.7 9.9916.3
Coμg g−1 240 35  55 150
Niμg g−1NA604 8281890
Cuμg g−1 12 11 8.2 8.2
Znμg g−1 105154  88 307
Gaμg g−1  4.2 1.9 3.5 3.0
Asμg g−1  1.0 1.0≤0.3 1.1
Seμg g−1  1.0 1.5 0.9 0.8
Rbng g−1 255 19≤15  37
Srng g−1 5516131900 104
Yng g−1 6103802600 180
Zrng g−1 470100 ± 201500 100
Nbng g−1 180 20  30  30
Mong g−11600110 300 300
Rung g−1 300≤1340 380NA
Agng g−13040 ± 10  20  30
Cdng g−1NA 35NANA
Inng g−1NA≤3NANA
Pdng g−1 140 30  60  60
Snng g−1 21030 ± 10  20  60
Sbng g−1 100≤25  20≤18
Teng g−1 120140  70 120
Csng g−1  19.2 0.4 2.0 2.5
Bang g−1 24743 ± 12 421 273
Lang g−1 27 5.1  50 4.2
Ceng g−1 58 17 130  10
Prng g−1  9.5 3.9  35 1.5
Ndng g−1 48 27 254 6.7
Smng g−1 24 16 160 4.0
Eung g−1  5.6 5.1  22 0.8
Gdng g−1 43 29 270 8.6
Tbng g−1 10 7.1  62 2.5
Dyng g−1 66 50 34717.9
Hong g−1 19 13  90 5.8
Erng g−1 65 42 259  21
Tmng g−1 13 8.1  42 4.7
Ybng g−1 81 54 256  35
Lung g−1 18 12  47 8.1
Hfng g−1 14  5  53   3
Tang g−1  0.5 0.08 0.6 0.1
Wng g−1 30011 ± 5  35  24
Reng g−1  65  5   7  48
Irng g−1 700 50  90 340
Ptng g−11000 70 100 500
Tlng g−1NA≤9NANA
Bing g−1≤2≤86NA
Thng g−
Ung g−

The majority of the elements we have analyzed agree well with prior available literature values for ureilites. In the case of several elements generally analyzed by instrumental neutron activation analysis (INAA) discussed next, either upper limits or no values were available for comparison. Of particular interest are the following results.

Ba––Literature values for Ba in ureilites rarely reach the values found in fragments #7, #15, and #47, whereas our value for fragment #4 is more in line with literature data that generally reports <30 ng g−1 with occasional excursions to the values we report for Almahata Sitta. We should note that our Ba values for USGS standard reference materials are actually 16–36% lower than literature values (see Appendix). However, we can find no reason to view our Ba values with more than the usual suspicion since the values were above limits of detection.

Ti––Previously reported Ti values determined by INAA are mostly upper limits (Warren et al. 2006; and references therein), which are typically reported to be well below 800 μg g−1, which is in good agreement with our values that cluster around 200 μg g−1, with a high value of 676 μg g−1 for fragment #15 (Table 1).

K––The moderately volatile element K is typically present in low concentrations (<0.1 mg g−1) in ureilites (Warren et al. 2006, and references therein) and Almahata Sitta is no exception. In the three fragments where it was determined, K concentrations were below our limit of detection (0.03 mg g−1), which is in good agreement with prior findings.

Hf––Literature reports (Warren et al. 2006; and references therein) generally show that Hf contents in ureilites rarely reach greater than approximately 50 ng g−1, since upper limits for this element are again typically reported and our results for Almahata Sitta concur with typical values well below this (Table 1).

Despite the inherent variability of the volatile trace elements in differentiated meteorites our Bi, Zn, Te, Se, Cd, In, and Tl results accord well with data from ureilites previously analyzed by either radiochemical neutron activation analyses (RNAA) or INAA (Binz et al. 1975; Wang and Lipschutz 1995; Warren et al. 2006 and references therein). Values for elements for which we were able to determine absolute concentrations (Bi, Zn, Te, Se, and Cd) or upper limits (Tl and In) are both bounded by literature values for ureilites.

As we will discuss further, with few exceptions, the elemental composition of Almahata Sitta is generally within prior values reported for ureilites. Deviations or compositional peculiarities are discussed in detail in the Discussion section.


Rare Earth Elements

We quantified all 14 extant rare earth elements (REE) in each fragment of Almahata Sitta. Results are shown in Fig. 1. Overall, the REE abundances are low (<0.5 × CI), except for results for fragment #15. The generally low concentrations are consistent with a ureilitic composition (Fig. 1). However, many ureilites possess a V-shaped REE pattern where La through Sm (the light REE or LREE) are sequentially depleted, have a negative Europium anomaly, followed by a progressive enrichment moving stepwise from Gd to Lu (heavy REE, HREE). REE in each of the four fragments of Almahata Sitta analyzed for this work do not show this V-shaped REE pattern, but rather are LREE-depleted (Fig. 1), but none of the depletions retreat below 0.01 × CI. It has been shown with analyses of multiple fragments of Roosevelt County 027 that V-shaped patterns can exist within the same rock as LREE depleted patterns (Goodrich et al. 1987), but the evidence so far would seem to lean toward Almahata Sitta possessing solely a LREE-depleted pattern. The slope of HREE elements (Gd > Lu) is steeper in the case of the highest LREE depletion (Fig. 1). In contrast, the relative magnitude of the Eu anomaly does not correlate with degree of depletion. LREE elements (Sm) show very different patterns in each sample, with coarse-grained samples (#4, #15, and #47) having the steeper LREE slopes.

Figure 1.

 REE abundances in four fragments of Almahata Sitta. REE patterns of Almahata Sitta show the depleted patterns associated with ureilites but also possess relatively depleted LREE patterns. Although many ureilites possess V-shaped REE patterns where La-Sm are progressively depleted and HREE (Gd-Lu) progressively enriched, this feature is absent in Almahata Sitta, which shows a progressive enrichment from La-Lu. CI (Orgueil) reference values for all figures are from Friedrich et al. (2002) and Anders and Grevesse (1989).

Polymict ureilites such as EET 83309, EET 87720, North Haig, and Nilpena are known to have REE contents that are generally higher than other ureilites, and Jenniskens et al. (2009) took the pattern for fragment #7, which does not fall below 0.1 × CI, to support a polymict chemical affinity for Almahata Sitta. With the addition of our new data, especially for that of fragment #4, which was sampled at the same mg scale that most INAA REE data were gathered, this may not be the whole story. Our results for fragments #4 and #47, representing both the compact pyroxene-rich and olivine-dominated lithologies, show severely depleted LREE patterns and HREE contents barely manage to break 0.2 × CI.

Since the majority of ureilites are finds, many with extended Antarctic or hot-desert residence times, our data on the freshly fallen Almahata Sitta points to terrestrial contamination as a source of the V-shaped REE patterns found in many ureilites. Since our primary concern was to obtain bulk compositional data, we performed no leaching experiments like those of Spitz and Boynton (1991). Such experiments on Almahata Sitta may yield information about the carrier and origin of the LREE phase in the ureilites and should be a priority for future experiments.


Figure 2 shows CI-normalized and, to remove the effect of volatiles (such as C) from the comparison, Mg-normalized abundances for several representative lithophile elements in Almahata Sitta, several polymict ureilites, and ranges of abundances within ureilites as a group. The general abundances are consistent with an ultramafic mineralogy being relatively enriched in Sc V, Cr, Mn, and despite its volatile nature, Zn (see Volatile Trace Elements section below).

Figure 2.

 CI- and Mg-normalized lithophile element abundances in four fragments of Almahata Sitta and, where available, polymict ureilites. Ranges of literature values are shown as horizontal lines. A mean of the Mg concentration (22.2%) of the other three fragments was used for normalization of fragment #7 data and a mean value for Mg in ureilites was used for normalization of literature Sr, Rb, and Cs rather than individual values since few determinations exist for both the element in question and the normalizing element in the same sample.

There are 20 identified polymict ureilites (Meteoritical Bulletin Database, 2009), and four (EET 83309, EET 87720, Nilpena, North Haig) of those have substantial compositional data available for comparison. Mittlefehldt et al. (1998) noted that polymict ureilites in general possess higher abundances of Al, Cs, K, Na, and Zn than average ureilites. The abundances of Cs and Zn in Almahata Sitta fragment #7 are all at the higher end of the ureilite composition range and this observation was a factor in the classification of Almahata Sitta as a polymict ureilite (Jenniskens et al. 2009). Here we find that other fragments of Almahata Sitta do not necessarily share the same enrichments with fragment #7: samples of fragments #4, #15, and #47 analyzed by us tend toward compositions of typical (monomict) ureilites.

We collected data for several other lithophile elements (Li, Y, Zr, Nb, Hf, Ta, Th, and U) that are not shown in Fig. 1 since little literature data were available for direct comparison. However, we can comment that the data omitted from Fig. 1 both reflect the ultramafic composition of ureilites and at the same time seem to parallel the compositional trends demonstrated by those elements shown in Figs. 1 and 2. For example, Y, Th, and U in fragment #15 are all higher than those found in the other fragments analyzed, as may be expected since REE are also enriched in that sample (Fig. 1).

Ratios of some important lithophile elements can shed light on the composition of Almahata Sitta. CI-normalized Ca/Al ratios in Almahata Sitta #4, #15, and #47 are respectively 4.3, 9.5, and 6.3. These values are known to range between 0.6 and 8.0 (Goodrich 1992). CI-normalized Mn/Mg ratios are respectively 0.73, 0.78, 0.70, which compare well with the typical value of approximately 0.7 in ureilites. CI-normalized Cr/Mg ratios in Almahata Sitta fragments are 0.61, 0.57, and 0.87. The values of fragment #15 are generally lower than those found in ureilites, which rarely have values lower than approximately 0.6 and, again, some degree of sampling error may account for this. However, overall, from the perspective of major and minor elements, Almahata Sitta is clearly a ureilite. Almahata Sitta does not in any way compositionally resemble Lewis Cliff (LEW) 88774, a thoroughly anomalous monomict ureilite (Chikami et al. 1997) that is enriched in Sc, Ca, V, and especially Cr.


Siderophile concentrations in ureilites are known to be inhomogeneous and quite variable within centimeter-distance subsamples of even the same meteorite (Rankenburg et al. 2008, and references therein). These variations have been known to be up to thousands of a percent, so some caution must be taken with interpreting bulk abundances from our small analyzed fragments. To compensate for sampling effects, we show our siderophile data as both CI- and Fe-normalized. Even so, we should point out that simply CI-normalized siderophiles and moderately siderophile elements in Almahata Sitta, typically fall within previously encountered values for ureilites.

Upon examining the data in Fig. 3, we can see that for especially Ir, Pt, Ni, and Co, siderophile contents parallel one another. That is, for fragments high in one element, other siderophiles are also higher in abundance. We attribute this to the fact that siderophiles in ureilites seem to derive from at least two components (Boynton et al. 1976; Rankenburg et al. 2008). One component is characterized by superchondritic abundances of refractory siderophile elements such as Ir and Pt, while the other component is closer to chondritic in composition. Aside from Ru, for which our uncertainty is larger than for other elements (Table 1), abundances of W through Ga (n = 10, Fig. 3) all reflect similar contributions from these two components. However, as we increase in volatility and decrease in siderophilic character with Ag, Sb, As, and Sn, the contribution from the second, more volatile component is more variable. Overall, our additional data for siderophiles supports and augments the data available for the multiple, but common components of siderophile elements in ureilites.

Figure 3.

 CI- and Fe-normalized siderophile and moderately siderophile element contents in four fragments of Almahata Sitta compared with ranges of compiled literature values for ureilites. Elements are arranged in order of putative volatility. Ranges of literature values are shown as horizontal lines. A mean of the Fe concentrations (12.3%) of the other three fragments was used for normalization of fragment #7 data, which seems atypically enriched in siderophiles as a result of a compositionally unrepresentative fragment. Generally, highly siderophile (e.g., Re, Ir, Pt, Ni, Co, Pd) elemental abundances parallel one another indicating a common reservoir of those elements in Almahata Sitta. The more volatile elements at the right show greater variability, probably reflecting a compositionally variable second chondrite-like complement of moderately siderophile elements.

Our values for Ag are higher than those previously reported for ureilites. The apparently high Ag values (Fig. 3) in our samples of Almahata Sitta may reflect a low maximum value for Ag literature data rather than a bias within our results since the mean (n = 22) Ag value for literature data is 18 ng g−1, while our mean value and 2σ for four Almahata Sitta fragments is 30 ± 12 ng g−1, barely coinciding with reported values. The origin of the discrepancy likely arises from a lack of ureilites that have quantified both Fe and Ag within the same analytical aliquot (n = 0) rather than simply for the same meteorite (n = 6). The latter is what is shown in Fig. 3, but we have indicated that these values are suspicious in Table 1. Although our Almahata Sitta values for Ag seem discordant, we point out that our Ag values for DNC-1 are actually lower than certified values (see Appendix).

Volatile Trace Elements

Mean concentrations of elements possessing greatest volatility in our 60 element suite are generally depleted in Almahata Sitta consistent with previous ureilite studies (Wang and Lipschutz 1995). The mean CI-normalized concentrations of four highly volatile trace elements for which we report absolute concentrations (Bi, Te, Se, and Cd) demonstrate a relatively flat trend with a mean value 0.050 ± 0.002. CI-normalized concentrations for Tl and In are <0.06 and <0.04, respectively. Zn, however, is an exception to this trend. Concentrations of Zn in our four fragments are significantly enriched with respect to the other highly volatile elements. The concentration of Zn in fragment #47 approaches CI concentrations. The mean CI-normalized Zn concentration for our four fragments is 0.53 ± 0.32. High Zn concentrations have been attributed to its proxying for Fe2+in olivine (Goodrich 1992), in accordance with our higher Zn values for the olivine-dominated fragment #47. Together, these results show that any evidence for nebular fractionation is obscured by later parent asteroid igneous processes.

Petrographic and Isotopic Correlations with Trace Elements

Although we investigated the compositions of each of the lithologies represented by the current range of Almahata Sitta samples, the size of sample aliquots chemically analyzed was small. Nevertheless, we can preliminarily examine how the observed variations in composition may potentially relate to oxygen isotopic composition and gross lithology texture. Rumble et al. (2010) found that Almahata Sitta fragments #4, #15, #44, #47, and #49 seem to cluster with respect to Δ17O and hypothesized that this subgroup of fragments may record a localized volume within the Almahata Sitta parent asteroid. Although our samples were small, potentially introducing sample size bias effects, most Mg- or Fe-normalized respective lithophile and siderophile abundances of these samples do not cluster more closely together with themselves than with fragment #7 (e.g., Figs. 2 and 3). We did not have an opportunity to analyze any materials other than fragment #7 that lie outside the cluster of Δ17O isotopic ranges defined by the subgroup of fragments #4, #15, #44, and #47 and this may be of interest for further study.

Examining correlations with lithology, we see no obvious compositional correlations with petrographic properties other than those that may be related to sampling. While fragments #4 and #15 were of the compact pyroxene-rich lithology (Zolensky et al. 2010), #47 seems comprised of the olivine-dominated compact lithology (Sandford et al. 2010), and fragment #7 is finer grained (Zolensky et al. 2010), we see no correlating chemical properties in terms of lithophile elements including the REE. Fragment #7 does stand out with respect to siderophile elements, but as discussed, we attributed this to an inhomogeneous sampling of the material. The difficulty in sampling may be related to its fine-grained nature. Finally, high values of Rb and Cs were originally regarded as supporting evidence for the polymict nature of Almahata Sitta by Jenniskens et al. (2009), but while it is true that polymict ureilites tend to be enriched in incompatible lithophile elements, the inverse is not necessarily true. We refer the reader to other works in this volume to ascertain further correlations, if any.


Almahata Sitta has a composition consistent with other ureilites. REE patterns in Almahata Sitta are LREE-depleted, with negative Eu anomalies and HREE enrichments. The abundances of REE only reach CI values in the case of our sample of fragment #15. The LREE enriched V-shaped pattern found in many ureilites is absent within our samples of Almahata Sitta. Since Almahata Sitta was collected shortly after its fall, these REE compositional results indicate that the V-shaped REE pattern found in many ureilites may be a result of terrestrial alteration. Of the fragments analyzed for this work, trace lithophile results from fragment #7 seem to more closely resemble those of other polymict ureilites, but the abundances for the same elements in fragments #4, #15, and #47 seem to most resemble typical monomict ureilites. However, we point out that chemical identification of a mono or polymict chemical connection is tenuous at best. Finally, we have seen that CI- and Fe-normalized siderophile abundances in Almahata Sitta are consistent with a two-component model of petrogenesis, with highly siderophile elements displaying parallel enrichments across all fragments and less siderophilic elements present in more variable abundances.

Acknowledgments— We thank the students and staff of the University of Khartoum for their efforts in recovering these meteorites and to kindly make samples available for analysis. Portions of this work were supported by NASA under the Planetary Geology and Geophysics program through grant NNX09AD92G to J. M. F. and Indiana State University Promising Scholars Award to S. F. W. D. R. thanks NASA’s Cosmochemistry program for support through grant NNX07AI48G. J. T. gratefully thanks the Clare Boothe Luce Program for support. P. J. is supported by a grant from the NASA Planetary Astronomy program. Reviews by Prof. M. Ebihara and Dr. D. Mittlefehldt as well as assistance from Dr. A. Yamaguchi were much appreciated.

Editorial Handling— Dr. Akira Yamaguchi


To examine the precision and accuracy of our analytical techniques on ultramafic rocks such as ureilites, we analyzed the terrestrial mafic DNC-1 and ultramafic DTS-2 (sometimes listed as DTS-2b) USGS standard reference materials. DTS-2 possesses REE values comparable to those found in Almahata Sitta. Results and comparisons with literature values are presented in Table 2. Where possible, we compare our results with those of the USGS certificates of analysis, but in many cases, compiled values are more complete and recent. Particularly for DTS-2, few literature values are available for comparison for the suite of elements that we present data for here (Table 2).

Table 2.   Bulk composition of mafic and ultramafic USGS standards DNC-1 and DTS-2 and comparison with literature values.
ElementUnitDNC-1DNC-1 LITRef.DTS-2DTS-2 LITRef.
  1. aA weighted mean of these literature values was used for comparison.

  2. A: USGS Certificate of Analysis North Carolina Dolerite, DNC-1.

  3. B: Hu Z. and Gao S. 2008. Upper crustal abundances of trace elements: A revision and update. Chemical Geology 253:205–221.

  4. C: Ila P. and Frey F. A. 2000. Trace element analysis of USGS standards AGV2, BCR2, BHVO2, DTS2 and GSP2 by INAA. Journal of Radioanalytical and Nuclear Chemistry 244:599–602.

  5. D: Raczek I., Stoll B., Hofmann A. W., and Jochum K. P. 2001. High-precision trace element data for the USGS reference materials BCR-1, BCR-2, BHVO-1, BHVO-2, AGV-1, AGV-2, DTS-1, DTS-2, GSP-1 and GSP-2 by ID-TIMS and MIC-SSMS. Geostandards Newsletter 25:77–86.

  6. E: Eggins S. M., Woodhead J. D., Kinsley L. P. J., Mortimer G. E., Sylvester P., McCulloch M. T., Hergt J. M., and Handler M. R. 1997. A simple method for the precise determination of ≥40 trace elements in geological samples by ICP-MS using enriched isotope internal standardization. Chemical Geology 134:311–326.

  7. F: Govindaraju K. 1994. 1994 compilation of working values and sample description for 383 geostandards. Geostandards Newsletter 18:1–158.

  8. G: USGS Certificate of Analysis Twin Sisters Mountain Dunite, DTS-2.

  9. H: Jochum K. P., Stoll B., Pfänder J. A., Seufert H. M., Flanz M., Maissenbacher P., and Hofmann M. 2001. Progress in multi-ion counting spark-source mass spectrometry (MIC-SSMS) for the analysis of geological samples. Fresenius Journal of Analytical Chemistry 370:647–653.

  10. I: Nakamura K. and Chang Q. 2007. Precise determination of ultralow (sub-ng/g) level rare earth elements in ultramafic rocks by quadrupole ICP-MS. Geostandards and Geoanalytical Research 31:185–197.

Liμg g−14.15.1E1.5 
Namg g−114.113.9A0.02960.0275C
Pmg g−10.2700.305A<0.01 
Kmg g−12.471.94A0.003280.00312D
Scμg g−1   3131.0E2.6   3G
Tiμg g−1 3590 2878E  31 
Vμg g−1  124  148E  23  22G
Crmg g−10.2650.270A14.715.5G
Mnmg g−11.241.16A0.8040.830G
Coμg g−1   69   55E 131 120G
Niμg g−1  249  247A36203780G
Cuμg g−1  132   96E2.9   3G
Znμg g−1   88   66E  41  45G
Gaμg g−116.7   15E0.65 
Asμg g−10.170.2F0.20 
Seμg g−10.550.2F0.05 
Rbng g−1 3270 4500E  1812.3D
Srng g−1119000145000E 413 534D
Yng g−12000018000E  38  49H
Zrng g−15100041000E 200 410H
Nbng g−1 1300 2000E  12  24H
Mong g−1  200  190B  20 
Rung g−1    7    9 
Agng g−1   90   27F  10 
Cdng g−1  133   85B  94 
Inng g−1   46   48B  15 
Pdng g−1  220   16F   5 
Snng g−1 1600 1470B 650 
Sbng g−1 1000  880B 580 600G
Teng g−1   30   22B  13 
Csng g−1  230  300E1.3 
Bang g−198800118000E1020016000G
Lang g−1 39603530E13.4  13D,Ia
Ceng g−1 91008190E  27  26D,Ia
Prng g−1 12001100E3.53.3D,Ia
Ndng g−1 51004860E  13  14D,Ia
Smng g−1 16001400E3.23.2D,Ia
Eung g−1  630 600E1.80.88D,Ia
Gdng g−1 22002000E3.73.4D,Ia
Tbng g−1  430 390E0.620.63D,Ia
Dyng g−1 24002760E3.74.4D,Ia
Hong g−1  630 650E1.21.3D,Ia
Erng g−1 19401900E4.85.1D,Ia
Tmng g−1  330 330F1.21.2D,Ia
Ybng g−1 19001970E9.710.2D,Ia
Lung g−1  360 309E2.52.2D,Ia
Hfng g−1 12001050E   6 
Tang g−1   50  98E0.2 
Wng g−1   50  56B   9 
Reng g−1    4 0.1 
Irng g−1   70.5F   3 
Ptng g−1  50   36F   2 
Tlng g−1   22   26B<0.2 
Bing g−1   10   11B4 ± 1 
Thng g−1  280  220E5.5 
Ung g−1   66   50E2.2 

Analytical Accuracy

For the major and minor (Na, Mg, Al, P, K, Ca, Cr, Mn, Fe, and Ni) and REE, our values rarely show deviations greater than 10% from literature values. Some exceptions include Dy and Lu, which are about 15% higher and lower, respectively. While our Eu value for DNC-1 is only 5% lower than accepted values, our Eu value for DTS-2 is half of that compiled from the literature. Perhaps this is due to a paucity of literature data available––only two REE analyses were found for DTS-2 in the literature. Our results for other lithophiles such as Sc, Ti, V, Cu, Sr, Y, Zr, Nb, Ba, Hf, Ta, Th, and U typically have larger deviations from literature values. On average, literature values agree with our own to within 20% for these elements, with the high field strength elements Zr, Nb, Ta showing the largest deviations (up to 50%), especially when compared with the DNC-1 dolerite.

Moderately volatile and thermally labile elements (Zn, Ga, As, Se, Rb, Cd, In, Sn, Sb, Te, Cs, Tl, and Bi) results likewise compare well with prior values with deviations rarely reaching 36% (Te with 30 ng g−1 literature versus 22 ng g−1 for our value for DNC-1). Our lower determined Se value for DNC-1 (0.2 μg g−1 versus 0.55 μg g−1 literature) is another discrepant value. Our thermally labile elements Cd, In, Tl, and Bi data agree extremely well with literature values despite their extremely low concentrations.

Very few analyses exist for siderophile elements in terrestrial ultramafic rocks. However, when compared with available literature values, our siderophile (Co, Mo, Ru, Ag, Pd, W, Re, Ir, and Pt) accuracy is best for Co, Mo, W, and Pt. Larger differences are apparent when considering Ag, Pd, and Pt (Table 2). No literature values are available for Re comparison. The single literature value for Pd in DNC-1, shown for comparison in Table 2, is different than ours by a factor of approximately 10 as is our value for Ir, with comparisons again based on the only literature value available.

Analytical Precision

Our analytical precision for replicate subsamples of the same fragment of Almahata Sitta is discussed in the text, but evaluating the reproducibility of our methodology on homogenized USGS standard powders gives further insight into the homogeneity of Almahata Sitta. For DNC-1 the mean relative standard deviation for all of our determined elements is 7.9%. The same for DTS-2 is 11.6%. The higher relative standard deviation for DTS-2 reflects the lower concentrations of trace elements within the standard. Since Almahata Sitta has trace element values that are more akin to DTS-2 than DNC-1, we feel the precision of DTS-2 is a good indicator for the reproducibility of our Almahata Sitta results in this work.