Chromium isotope anomaly in an impactite sample from the El'gygytgyn structure, Russia: Evidence for a ureilite projectile?


  • Julien Foriel,

    1. Department of Earth and Planetary Sciences and McDonnell Center for the Space Sciences, Washington University in St. Louis, St. Louis, Missouri, USA
    Search for more papers by this author
  • Frederic Moynier,

    1. Department of Earth and Planetary Sciences and McDonnell Center for the Space Sciences, Washington University in St. Louis, St. Louis, Missouri, USA
    Search for more papers by this author
  • Toni Schulz,

    1. Department of Lithospheric Research, University of Vienna, Vienna, Austria
    Search for more papers by this author
  • Christian Koeberl

    Corresponding author
    1. Department of Lithospheric Research, University of Vienna, Vienna, Austria
    2. Natural History Museum, Vienna, Austria
    • Department of Earth and Planetary Sciences and McDonnell Center for the Space Sciences, Washington University in St. Louis, St. Louis, Missouri, USA
    Search for more papers by this author

Corresponding author. E-mail:


The 3.6 Ma, 18-km-diameter El'gygytgyn impact structure (Arctic Russia) is unique among the currently known terrestrial impact craters in that it is the only one that was formed in acid volcanic rocks. Previous analyses of impactites from El'gygytgyn showed minor enrichments of the siderophile elements, including Ir, which, together with distinct Cr enrichments, gave rise to speculation that an achondritic projectile was involved. We studied the major and trace element composition in samples from the new ICDP drill core obtained near the center of the structure, as well as the chromium isotopic composition of an impact glass sample collected on the surface. Several suevitic breccias from the upper part of the suevite sequence in the drill core show higher Cr and Ni contents compared with felsic volcanic rocks in the lower part of the core and from surface samples. However, it is difficult to unambiguously establish a meteoritic component from trace element data, as input from (rare) mafic target rocks is a possibility. In contrast, the Cr isotopic composition of the impact glass sample yielded a nonterrestrial ε54Cr value of −0.72 ± 0.31 (2 std. err.). This negative ε54Cr is different from known carbonaceous chondrite values (ε54Cr of +0.95 to +1.65), but is nearly identical to reported values for ureilites (approximately −0.77). The value is, however, also within analytical error of eucrites (approximately −0.38) and ordinary chondrites (approximately −0.42). Given the chemical signatures found in previous analyses of El'gytgytgyn impactites and the similarity of our Cr isotopic data to ureilites, we suggest that the impacting asteroid could have been an F-type asteroid of mixed composition, similar to the recent Almahata Sitta fall in Sudan.

Introduction and Geological Background

The 18-km-diameter El'gygytgyn impact structure is located on the Chukotka Peninsula in Arctic Russia (67º30′N, 172º34′E). The circular structure, now largely filled by the about 12-km-diameter Lake El'gygytgyn, was confirmed as of impact origin by the discovery of shock metamorphic effects in the crater rocks by Gurov et al. (1978, 1979); see also Gurov and Gurova (1979, 1991). Details of earlier work are summarized in Gurov and Koeberl (2004) and Gurov et al. (2005, 2007). El'gygytgyn is a well preserved and young impact structure, somewhat comparable to Bosumtwi in Ghana (see Koeberl et al. 2007a). The crater-forming impact was dated at a mid-Pliocene age of approximately 3.6 Ma (3.58 ± 0.04 Ma according to the 40Ar/39Ar dating by Layer 2000). What makes El'gygytgyn unique among the currently known terrestrial impact craters is that it is the only one that was formed in acidic volcanic rocks, thus giving the opportunity to study shock metamorphic effects in these rock types. Target rocks at the El'gygytgyn area are dominated by the Late Cretaceous volcanic suites of the Okhotsk-Chukotka Volcanic Belt and include volcanic series of lava, tuff, and ignimbrites of rhyolitic and dacitic composition, andesite, and andesitic tuffs (cf. Gurov et al. 2005, 2007).

Shock metamorphosed rocks and impact melt rocks occur in the El'gygytgyn structure as redeposited materials in lacustrine terraces inside the crater and, locally, in terraces along little streams on the outer slope of the crater rim. Two main types of impact melt rocks and glasses have been found in these locations and include aerodynamically shaped glass bombs and shock metamorphosed breccias (e.g., Gurov and Koeberl 2004; Pittarello and Koeberl 2013a). Gurov et al. (2005) proposed that such impact melt rocks originated from the ejecta blanket, which had been completely eroded and slumped off the rim. The impactites are generally fresh and neither display significant postimpact hydrothermal alteration nor alteration due to weathering (Gurov et al. 2005). As there are melt rocks in the target as part of the volcanic (rhyolitic) sequence, the distinction between volcanic melts and impact melts can be difficult (Pittarello and Koeberl 2013b; Raschke et al. 2013a). However, the impact glasses are relatively fresh compared with altered and recrystallized melts in the volcanics deeper in the sequence, and they can contain clasts with diaplectic glass or other indications of shock metamorphism.

An important question at any impact structure regards the nature of the impacting body. The detection and verification of an extraterrestrial component in impact-derived melt rocks or breccias can be of diagnostic value to provide confirming evidence for an impact origin of a geological structure (see Koeberl 1998, 2007; Koeberl et al. 2012). Generally, a very small amount of meteoritic melt or vapor is mixed with a much larger quantity of target rock vapor and melt, and this mixture later is incorporated into impact melt rocks or melt breccias, suevite, or impact glass. In most cases, the amount of extraterrestrial material within these impactite lithologies is much less than one percent by weight. Detecting such small amounts of meteoritic matter is very difficult and only elements that have high abundances in meteorites but correspondingly low abundances in terrestrial crustal rocks (e.g., siderophile elements such as Ni and Cr, and the platinum group elements [PGEs]) are used in such studies (cf. Moynier et al. 2009; Koeberl et al. 2012). Distinctly higher siderophile element contents in impact melts, compared with target rock abundances, can be indicative of the presence of either a chondritic or an iron meteoritic component (e.g., Palme et al. 1978; McDonald et al. 2001). Complications may arise: (1) because meteorites have a range of compositions within each class and some are better constrained than others (see McDonald 2002); (2) if the target rocks have variable siderophile element concentrations (McDonald et al. 2007); or (3) if siderophile element concentrations in the impactites are very low. Furthermore, the contribution of the target rock (the so-called indigenous component) to the composition of impactites can only be understood if either a well constrained mixing relationship exists between the impactor and the target rocks that produces a reliable regression line and a lower intercept that reflects the average PGE concentration in the target rocks (e.g., McDonald et al. 2001; Tagle and Claeys 2005), or all contributing target rocks have been identified and their relative contributions to the melt mixture are known—something that is very difficult to achieve in practice. This is part of the reason why the nature of the projectiles has not been identified for most of the known impact structures on Earth. Lack of preservation or exposure of melt rocks and breccias is another reason.

However, somewhat more specific than elemental abundances are isotopic indicators for a meteoritic component. Of these, the first one to be used with success was the Os isotopic system (see review by Koeberl and Shirey 1997). Here, the different isotopic compositions and significantly different abundances of the element Os are used to fingerprint a meteoritic component. This system is very sensitive but allows only detecting the presence of a chondritic or iron meteorite component, but without being specific to the meteorite type, and achondritic signatures cannot be identified. In contrast, the Cr isotope system is less sensitive, but offers selectivity between meteorite types, including achondrites (see e.g., Koeberl 2007; Koeberl et al. 2007b). Both methods are based on the observation that the isotopic compositions of the elements Os and Cr, respectively, are different between most meteorites and terrestrial rocks and that these differences are sufficiently large to permit detection of relatively small amounts of meteoritic Os or Cr present in the impact rock. The best candidate for detection of a meteoritic material are impactites in the form of melt rocks or breccias, either from a coherent impact melt body or as melt fragments in suevitic breccias, or ejecta such as impact glasses, all of which can incorporate impactor material (as solid, melt, or gas).

El'gygytgyn was recently the subject of a deep drilling project, supported by the International Continental Scientific Drilling Program (ICDP). Drilling was performed from late 2008 to mid 2009 (Koeberl et al. 2009; Melles et al. 2011). The goals of this project were an investigation of the Arctic paleoclimate evolution during the past 3.6 Ma, and a study of the impact rocks underneath the lake sediments to determine, inter alia, shock metamorphic effects in the siliceous volcanic rocks, or to try and determine the composition of the impactor. One deep borehole (D1c) was drilled into the impact lithologies beneath the lacustrine sediments. The borehole was located northeast of the lake center, on the flank of the central uplift, to a maximum depth of 517 m below lake bottom (b.l.b.). Details of the drilling project and the recovery of the drill core are given in Koeberl et al. (2013), and details of the lithologies, petrography, and geochemistry of the core rocks are given, e.g., in Pittarello et al. (2013) and Raschke et al. (2013).

Meteoritic Component at El'gygytgyn—Previous Work

An early study by Val'ter et al. (1982) of the geochemical composition of impactites at El'gygytgyn showed that the impactites are enriched in elements such as Ni, Co, and Cr compared with the contents in the target rocks. These authors attempted to measure the contents of Ir in the impactites as well, but found them to be below the detection limit of their method (<2 ppb). They interpreted the geochemical data of the impactites to reflect admixture of meteoritic material of possible ureilitic composition. In a slightly later investigation, Kapustkina et al. (1985) succeeded in measuring the Ir contents of target rocks and impact glass and found very low values, but slight enrichments of Ir compared with target rock values in some impact glasses. They concluded that these values cannot be explained by selective vaporization and might be in agreement with the suggestion of an achondritic projectile by Val'ter et al. (1982).

Two more recent studies have tried to shed more light on the composition of the impactor at El'gygytgyn by comparing the composition of impactites with those of the target rocks. Both Gurov and Koeberl (2004) and Goderis et al. (2013) showed that the composition of the impactites varies little from that of the basement rocks in terms of major and most trace elements. Melt glasses and bombs analyzed by Gurov and Koeberl (2004) show minor enrichments in Cr (20–46 ppm) and Ir (24–110 ppt) compared to a partially melted target rhyolite sample (Cr = 8 ppm, Ir not detected). Using gamma–gamma coincidence spectrometry, Gurov and Koeberl (2004) found iridium contents in impact melt rocks and glasses that ranged from 24 ppt in a glass bomb to 74 and 111 ppt in impact melt rocks, which are not particularly high values compared with those found in impactites of other impact craters (e.g., Koeberl 2007). This led Gurov and Koeberl (2004) to suggest that an achondrite could be responsible, but they also cautioned that no unambiguous conclusions can be drawn regarding the presence and nature of a meteoritic component.

Moreover, Goderis et al. (2013) analyzed three impact glass/melt rock samples from the crater surface, as well as 17 drill core samples from the El'gygytgyn ICDP hole D1C (see also Wittmann et al. 2013) for their geochemical composition and found coupled enrichments of Cr, Ni, and Co for some of the samples; they also analyzed eight drill core and two impact melt rock samples (collected at the crater surface) for their PGE contents, and four drill core samples and one glass surface sample for their Os isotopic compositions. They also found that some samples, especially those near the top of the impactite sequence (the top of the possible fallback sequence), have slightly elevated contents of Ir (similar to values found by Gurov and Koeberl 2004) and some other PGEs, as well as lower 187Os/188Os ratios compared with crustal rocks and local target rocks (all of which indicate the presence of a meteoritic component in El'gygygtyn rocks). Goderis et al. (2013) also analyzed the composition of impact glass spherules and spherule fragments that were recovered from the reworked fallout deposits and from terrace outcrops of the Enmyvaam River approximately 10 km southeast of the crater center (see also Smirnov et al. 2011). Based on their siderophile element (including PGE) data and Os isotopic composition, these authors first considered an achondritic (ureilitic) projectile to be likely, but then concluded that their data provide a better fit for an ordinary chondrite projectile, mainly based on the spherule abundance data.

Although it must be noted that none of these studies came up with a conclusive finding, all these data mentioned above provide geochemical support for the presence of a meteoritic component in some El'gygygtyn impactites. The nature of the impactor, however, remained poorly constrained. To further constrain the possible meteoritic component in these impactites, we therefore have used a combined trace element and Cr isotopic study.

Chromium Isotope Method

Chromium isotopic anomalies have long been recognized in CAIs (Birck and Allègre 1984; Papanastassiou 1986) and in leachates from carbonaceous chondrites (Rotaru et al. 1992; Podosek et al. 1997), and were used by Shukolyukov and Lugmair (1998) to demonstrate the presence of extraterrestrial material in K/Pg boundary layer samples. Systematic 53Cr and 54Cr anomalies in different groups of meteorites were later identified by Shukolyukov and Lugmair (2006a), Trinquier et al. (2007, 2008), Qin et al. (2010b), and Moynier et al. (2009). For example, compared with the terrestrial standard value, carbonaceous chondrites exhibit 54Cr excesses ranging from +0.8 to +1.6 ε-units, while 54Cr is depleted in ordinary chondrites (ε54Cr ≈ −0.4), eucrites (ε54Cr ≈ −0.7), and ureilites (ε54Cr ≈ −0.9) (see the 'Results and Discussion' section). Chromium-53 shows lesser variations with maximum enrichments around 0.7 ε in eucrites (see 'Results and Discussion'). These variations may be attributed to a heterogeneous distribution of different presolar carrier phases of anomalous Cr within the solar system.

Many meteorite types contain large concentrations of Cr (from 0.01 to approximately 1 wt%), whereas the Earth's upper continental crust typically contains around 90 ppm Cr (e.g., Rudnick and Gao 2003). Therefore, even a very small contribution of bulk meteoritic material to impactites is sufficient to result in a chromium pool that contains enough extraterrestrial Cr, so that ε54Cr is measurably different from the terrestrial value. For example, a 0.5% inclusion of impactor material containing 4000 ppm Cr in a 90 ppm Cr target rock, will lead to a total chromium pool of which about 20% of the Cr is extraterrestrial. The chromium isotopic composition can therefore be used to recognize extraterrestrial material and identify the type of the impactor. For discussions on how to apply this method to impact craters, see Koeberl et al. (2002, 2007b).

Samples and Methods

The El'gygytgyn drill core samples (for which trace element contents are discussed below) are described in detail by Pittarello et al. (2013); this publication also gives details on which samples were analyzed and by which methods, so that we refrain from duplicating this information here. In addition, chromium isotopic analysis was performed on impact glass El'gygytgyn G4 collected on the crater surface. This sample is similar to other glass bombs described in Gurov and Koeberl (2004) and its petrography and geochemistry is described in Gurov et al. (2005). Analogous to those glasses of the same group that were analyzed by Gurov and Koeberl (2004, their table 4) and Gurov et al. (2005, their table 5), G4 is inferred to contain slight Cr, Ni, and Ir enrichments above target rock levels, and is one of the El'gygytgyn samples in which a chemical contribution of the impactor—as suggested by the Ir content—might be present. Gurov et al. (2005) noted that sample G4 is a pumice-like impact glass of shock stage IV, the second-highest stage of Gurov et al. (2005), which is characterized by complete shock melting of volcanic rocks and tuffs. The sample contains diaplectic glass and few small crystalline clasts in a mostly isotropic and homogeneous, slightly altered, glassy groundmass that contains some small birefringent needles as indication of incipient crystallization. The glass is mostly gray in color, but contains several centimeter-sized schlieren of colorless glass that could represent original silica-rich areas. The fabric of this sample is that of a laminated tuff, with a few darker bands in the glass. The rare crystalline clasts all have very low birefringence. Some resemble toasted quartz, and some unshocked relics of plagioclase are also present. Dark-brown elongated patches are probably relics after a mafic mineral such as biotite.

A fragment of El'gygytgyn G4 was ground into a fine powder with an agate mortar and pestle and homogenized; 1.04 g was dissolved in an approximately 5:1 concentrated HF:HNO3 mixture and heated for several days and, after drying, redissolved in concentrated HCl to remove fluorides. A basalt geostandard (BHVO-2) and a sample of the K/Pg boundary at Gams (Austria) were processed in parallel. Multicollector ICP-MS analysis requires the complete removal of titanium, vanadium, and iron before analysis of Cr, as these elements create isobaric interferences of 50Ti and 50V on 50Cr and 54Fe on 54Cr.

The Cr chemical purification is based on methods previously published for Cr-rich meteorites (e.g., Birck and Allègre 1985; Yin et al. 2007) and had to be adapted for the larger sample amount (1 g) required in a low Cr sample (see details in Table 1). The acid-sample solution was eluted in a cationic resin column in 1N HCl to remove most matrix elements and most of the iron and in HNO3, and in HF and HCl for a more complete removal of Ti, V, as well as the remaining Fe. The Cr recovery-yield at each step was >90%. Imperfect Cr recovery is not problematic because any stable isotope fractionation in the sample or induced by the chemical processing was corrected when calculating the 54Cr anomaly. Multiple elutions, typically two in each column, are required to reduce the level of interfering elements sufficiently. For example, reliable 54Cr data are obtained when Cr/Fe in the analyzed solution is approximately >3000, as opposed to Cr/Fe approximately 0.001 in the starting rock sample.

Table 1. Detailed Cr chemical purification methods
 Eluted element(s)a
  1. a

    Elution fraction estimated from test runs with elemental standard solutions loaded.

Steps 1 & 2. Large column
Resin: AG50X8 mesh 200–400, vol. = 1.6 mL 
Resin conditioning: H2O, 40 mL 
Sample loaded in HCl 1N solution, 6 mL 
Rinse, HCl 1N, 6 mL>90% Cr, 50–75% Ti and V, <0.2% Fe
Waste, HCl 6N>99% Fe, 25–50% Ti and V
Steps 3 & 4. Small column
Resin: AG50X8 mesh 200–400, vol. = 1.0 mL 
Resin conditioning: H2O, 40 mL 
Sample loaded in diluted HNO3, 3 mL 
Rinse, HF 0.5N, 5 mL~30% Ti, ~10% V, ~40% Fe
Rinse, HCl 1N, 16 mL~70% Ti, ~85% V, ~50% Fe
Rinse, HCl 2N, 2 mL~1% Ti, ~5%V, 10–20% Fe
Rinse, HCl 2N, 10 mL>90% Cr, <1% Ti, <5% V, <0.1% Fe

Isotopic measurements were performed at Washington University, using a Neptune Plus multicollector ICP-MS. Samples were redissolved in a 0.1N HNO3 solution and were introduced into the instrument through an Apex desolvator and a 100 μL/min nebulizer. To discriminate against the interference of argon molecules (argon is the carrier gas) at masses 54 (40Ar14N) and 56 (40Ar16O), the medium resolution mode of the Neptune was used with detector cups at argon interfering masses aligned on the flat 54Cr and 56Fe shoulder peaks. Minute amounts of Ti, V, and Fe remaining after the separation chemistry or contained in acid blanks need to be monitored and corrected for their interferences with 50Cr and 54Cr. Consequently, detectors were aligned as such: L3: mass 49 (49Ti), L2: mass 50 (50Cr), L1: mass 51 (51V), central cup: mass 52 (52Cr), H1: mass 53 (53Cr), H2: mass 54 (54Cr), H4: mass 56 (56Fe). Each analysis consisted of 30 acquisitions of 8 seconds each. Instrumental drift was corrected by using standard bracketing. Results are reported as part per 10000 (εXCr = [XCr/52Crin sample/XCr/52Crin standard−1] × 10000) and stable isotope fractionation is corrected using the measured 50Cr/52Cr and an exponential law.

Results and Discussion

The Case of a Meteoritic Component—Trace Element Correlations

The dominant volcanic series of the El'gygytgyn region, as well as most rocks from the drill core investigated in this study, are rhyolitic in composition. As noted above, such a felsic and homogeneous target lithology is unique among known terrestrial impact structures (Gurov et al. 2005) and potentially provides a good possibility for the search of an extraterrestrial component in impactites. Rhyolites have generally low concentrations of Ni, Cr, and Ir, whereas the abundances of these elements are orders of magnitude higher in meteorites. To search for traces of meteoritic components in the suevite from the uppermost layer of the El'gygytgyn drill core (between approximately 317 and 390 m b.l.b.), we compared concentrations of possible contaminant elements, typically common in impacting bodies, in the suevite with geochemical data from unshocked felsic rocks from the environs of the crater.

Compared with chondrites and iron meteorites, achondritic projectiles are chemically inconspicuous. A chemical identification of an extraterrestrial signature in suevites from the El'gygytgyn drill core could therefore be challenging. While generally low in siderophiles, achondrites exhibit concentrations of Cr up to 6000 ppm, of Co up to 120 ppm, and of Ni up to 1700 ppm. In all cases, ureilites exhibit the highest enrichments of these elements in all achondrite types (e.g., Spitz and Boynton 1991; Mittlefehldt 2003). Thus, the concentrations of Cr, Co, and Ni could provide an indication for the search of extraterrestrial signatures. Achondritic Cr, Co, and Ni concentrations would provide the highest enrichments compared with the average composition of the upper continental crust (UCC; 92 ppm, 17.3 ppm, and 47 ppm for Cr, Co, and Ni, respectively; Rudnick and Gao 2003) and the rhyolites in the El'gygytgyn area. In a first step, we, therefore tried to see if enrichments of some siderophile elements are also present in the new drill core samples, and if the conclusions of Val'ter et al. (1982) could be repeated for this material. Here, we discuss the new analyses obtained on samples from the El'gygytgyn ICDP hole 1C described by Pittarello et al. (2013).

In Fig. 1, the Cr, Co, and Ni concentrations in different types of meteorites and impactites are shown in comparison with the ranges of concentrations for different lithologies from the El'gygytgyn crater (see also Table 2). Felsic target rocks collected from outcrops in the vicinity of the crater, as well as ignimbrites from the lowermost layer of the drill core (between approximately 421 and 517 m b.l.b.), define upper concentrations of approximately 26 ppm Cr, 9 ppm Co, and 16 ppm Ni, whereas the suevites from Pittarello et al. (2013) range up to approximately 53 ppm Cr, 8 ppm Co, and 24 ppm Ni. We therefore defined those samples of suevite that exhibit higher Cr and Ni contents compared with felsic target rocks and ignimbrites as “enriched suevite” (see caption to Fig. 1 for details). However, some mafic rocks occur in the intermediate layer of the drill core (between approximately 390 and 421 m b.l.b.), as well as among the target rocks, and those were also analyzed by Pittarello et al. (2013). Their Cr, Co, and Ni contents exceed the respective concentrations of “enriched suevite” by more than one order of magnitude (see Fig. 1). El'gygytgyn impactites discussed by Gurov and Koeberl (2004) and Val'ter et al. (1982), presumably carrying meteoritic signatures, exhibit Cr, Co, and Ni concentrations comparable to the highest values measured in the suevite from this study (Fig. 1). In contrast, data for spherules collected from the top of the drill core in presumed fallback breccias and from a terrace deposit along the Enmyvaam River, the outflow of the crater lake, exhibit even higher Cr, Co, and Ni contents (Goderis et al. 2013; Wittmann et al. 2013). According to Goderis et al. (2013), the Ni/Cr and Cr/Co ratios for these spherules are within the compositional ranges of certain chondrites and primitive achondrites, and their Ni/Co ratios are intermediate to the ratios for ureilites and chondrites.

Table 2. Comparison of average Cr, Co, and Ni contents (in ppm) as well as Ni/Cr, Ni/Co, and Co/Cr ratios for meteorites and rocks from the El'gygytgyn crater
  1. Also shown are mixing calculations for a variety of target rocks (from the vicinity of the crater; data from Gurov and Koeberl 2004) with varying proportions of (1) basalt, (2) ureilite, and (3) LL-chondrite.

  2. a

    Basalts from the El'gygytgyn drill core (sample data from Pittarello et al. 2013).

  3. b

    Average data for suevites with enhanced Cr contents (samples 98Q2-W3-7 and 112Q7-W4-7 of Pittarello et al. 2013).

  4. c

    Data for impactites from Val'ter et al. (1982) as well as Gurov and Koeberl (2004).

  5. The mixing calculations show that our data for high-Cr suevites cannot be explained by any contribution from LL-chondrites (the impact type preferred by Goderis et al. 2013). Instead, ureilitic, or basaltic contamination from the target can account for the observed elemental data. Data for meteorites are from Mittlefehldt (2003), Wasson and Kallemeyn (1988), and Lodders (2003).

Target rocks~14.0~3.1~11.8~0.8~3.8~0.22
Mixing calculations
Target rocks
+5% Basaltsa~42.1~5.0~20.3~0.5~3.9~0.12
+0.7% Ureilites~42.3~3.7~18.9~0.5~5.1~0.09
+0.75% LL-chondrite~41.5~6.7~91.2~2.2~13.6~0.16
Figure 1.

Comparison of Cr, Co, and Ni contents for drill core and target rocks from the El'gygytgyn crater. Also shown are ranges for literature values of various “impactites” investigated by Val'ter et al. (1982) and Gurov and Koeberl (2004) together with values for respective target rocks from surface exposures. Also shown are data for spherules from the El'gygytgyn area collected and analyzed by Adolph and Deutsch (2010). Dotted lines represent average values. Typical element contents of various meteoritic materials are from Palme et al. (1978), Wolf et al. (1983), Janssens et al. (1987), Spitz and Boynton (1991), and Lodders (2003).

In Figs. 2 and 3, the Ni/Cr and Co/Cr ratios of rocks from the drill core (from top to bottom: suevite, mafic rocks, and ignimbrite) and of felsic target rocks are compared with literature data for impactites and a variety of meteorites. As can be seen in Fig. 2, rocks from the uppermost drill core layer (suevite), while generally defining Ni/Cr ratios around 0.8, extend to Ni/Cr ratios of approximately 0.3 (samples 1 and 2). This is comparable to the average value for ureilites and mafic rocks (basalts from the drill core). Felsic target rocks define a trend with Ni/Cr ratios of around 0.8, similar to most suevites (see also Table 2). Figure 3 shows the Co-Cr space and a similar trend for all drill core samples, with the “enriched suevite” (samples 1 and 2) again plotting off the regression line (defined by all other felsic drill core and target rocks). However, mafic rocks again exhibit elemental ratios comparable or even lower than those defined by the “enriched suevites.” It is remarkable that all data for El'gygytgyn impactites taken from Val'ter et al. (1982) and Gurov and Koeberl (2004) in the vicinity of the crater and that might carry an achondritic (e.g., ureilitic) component, exactly mirror the trend defined by “enriched suevites” in both diagrams.

Figure 2.

Nickel versus Cr abundances, showing interelement correlations for samples from the drill core. The target (dashed line B) defines Ni/Cr ratios of approximately 0.8, whereas suevites from the drill core define average Ni/Cr ratios of about 2. The area defined by the dotted border represents impactites from Val'ter et al. (1982) (postulated to contain a meteoritic signal) and Gurov and Koeberl (2004). Notably, this area overlaps with the composition of the two “enriched suevites” from the drill core (defined by triangles 1 and 2, representing samples 98Q2-W3-7 and 112Q7-W4-7, as described in Pittarello et al. 2013). Trends defined by basalts from the drill core (dashed line C) and various meteoritic materials are shown for comparison (for references see caption to Fig. 1).

Figure 3.

Cobalt versus Cr abundances, showing interelement correlations for samples from the drill core. The target (dashed line A) defines average Co/Cr ratios of approximately 0.22, whereas suevites from the drill core define average Co/Cr ratios of about 0.6. The area marked by the dotted border represents impactites from Val'ter et al. (1982) (postulated to contain a meteoritic signal) and Gurov and Koeberl (2004). This area overlaps with the composition of the two “enriched suevites” from the drill core (defined by triangles 1 and 2, representing samples 98Q2-W3-7 and 112Q7-W4-7, as described in Pittarello et al. 2013). Trends defined by basalts from the drill core (dashed line C) and various meteoritic materials are shown for comparison (for references see caption to Fig. 1).

Obviously, contamination of dacitic and rhyolitic suevite with mafic material from the same volcanic sequence, as seen in the drill core and in the target samples, could potentially account for any departures in the Cr, Co, and Ni signatures, obscuring or even simulating an extraterrestrial component in the “enriched suevite” investigated in this study. An inspection of thin sections (cf. Pittarello et al. 2013) from the drill core samples with the higher Cr, Co, and Ni contents revealed that mafic enclaves (e.g., basalt clasts) are present in the “enriched suevite,” but are generally absent in the (non-enriched) suevite and all other rocks investigated here (including the impact glass analyzed for Cr isotopic composition). This observation could indicate that the signatures in the “enriched suevite” might result from contamination with basaltic rather than meteoritic material. Such contaminations can be explained by the consequences of impact fracturing and brecciation, which could lead to juxtapositions of lithologies from varying depths. In the case of basaltic contamination, average compositions suggest a mafic component of approximately 5% in the “enriched suevite” (Table 2). On the other hand, Goderis et al. (2013) found enrichments in the contents of Ir and other PGEs, as well as a meteoritic Os isotopic signature, in suevite samples from the top of the drill core (but not in a mafic sample from the lower part of the drill core).

Thus, these data make it difficult to fully accept the interpretation of earlier studies (e.g., Val'ter et al. 1982), who, based on Cr, Co, and Ni contents, postulated a ureilitic impactor. Such a ureilitic component is, however, still possible—see the discussion below. Such a meteoritic interpretation would require around 0.7% ureilitic material (Table 2). Figures 2 and 3 also show different trends for other meteorite groups. Whereas a eucritic impactor can be ruled out because of low Ni contents (Fig. 1 and Table 2), LL-chondrites (and all other chondrites) are unlikely, as our mixing calculations with these components (Table 2) fail to mirror our data for “enriched suevites.” The same results were obtained by Goderis et al. (2013), whose data show that any suevite samples from the drill core have Ni/Cr, Ni/Co, and Cr/Co ratios that are unlike both chondrites and ureilites, even though the topmost suevites in the drill core do show PGE enrichments (and a low Os isotopic ratio). Only the Cr, Co, and Ni ratios for the spherules reported by Goderis et al. (2013) and Wittmann et al. (2013) are close to chondritic values; however, no Ir or other PGE abundances (or Cr or Os isotopic compositions) have yet been measured in these spherules and, thus, it is not clear what these results really tell us.

The selective vaporization induced by impact-triggered heating and melting could have led to depletions in volatile elements (e.g., Ries event; Osinski 2004). Notably, the highly volatile element Ga is significantly depleted in the El'gygytgyn suevite compared with all other drill core and target rocks. Whereas the suevite exhibits Ga contents between 1 and 4 ppm, rocks from the intermediate layer and ignimbrite range from 1 to 18 ppm, and other target rocks from 7 to 51 ppm. Furthermore, the K and Na contents in suevites are lower compared with ignimbrites and most felsic rocks from the transitional layer; in general, most target rocks exhibit higher K and Na contents compared with suevites, but no significant depletions were observed for other volatile elements, such as Br or Pb.

Therefore, it appears that it is difficult and impossible to constrain the presence of a meteoritic component in El'gygytgyn drill core samples based only on the discussion of Cr, Co, and Ni abundance data. The lack of precise data for iridium contents in the various target rocks (including the mafic components) makes it difficult to argue whether the slight Ir enrichments in impactites found by Gurov and Koeberl (2004) could be caused by an extraterrestrial component or not. Goderis et al. (2013) found some elevated Ir abundances in the uppermost impact breccia (fallback?), but also had one sample from a stratigraphically lower position in the drill core with a slightly elevated Ir content, which they labeled as “lower polymict impact breccia”; but they also noted a lack of shock features in that sample. The lack of any other apparent anomalies of nonvolatile elements in suevite compared with the unshocked target rocks collected at the crater, as well as the unshocked to slightly shocked volcanic rocks from the drill core, does not allow presenting any persuasive evidence for a meteoritic component in the suevite from the El'gygytgyn drill core based solely on element abundance data.

Thus, the approach of using isotopic fingerprinting in the search for a meteoritic component in El'gygytgyn rocks seems to be the only reliable way forward. Goderis et al. (2013) report a possibly meteoritic Os isotopic signature. Our approach was to use the potentially more selective Cr isotopic method. In our analysis, the chromium concentration estimated from MC-ICP-MS data in impact glass sample G4 is 20 ppm, similar to the range observed for other El'gygygtyn glass bombs (17–46 ppm Cr, Gurov and Koeberl 2004; Gurov et al. 2005; Goderis et al. 2013). Because of this low concentration, only three measurements of the Cr isotopic composition of G4 could be achieved: ε53Cr = −0.10 ± 0.25 (2 std. err., n = 3) and ε54Cr = −0.72 ± 0.31. Replicated analyses of the BHVO-2 geostandard yielded ε53Cr and ε54Cr values that are typical for all terrestrial samples (ε53Cr = −0.10 ± 0.10 and ε54Cr = −0.03 ± 0.24, 2 std. err., n = 14). We also analyzed a Cretaceous/Paleogene (K/Pg) boundary sample (from Gams, Austria) and measured similar 53Cr and 54Cr enrichments to those reported by Trinquier et al. (2006) for K/Pg boundary samples at different locations (Table 3), which concurs with a carbonaceous chondrite impactor for this impact event. The agreement between our results for terrestrial and K/Pg samples and previously published data suggests that our results for the El'gygytgyn sample are accurate.

Table 3. Chromium isotopic data for El'gygytgyn impact glass, meteorites, and terrestrial samples
This studyε53Cr2 std. err.ε54Cr2 std. err.
K/Pg boundary Gams0.200.170.780.27
El'gygytgyn G4−0.100.25−0.720.31
Previous studiesε53Cr2saε54Cr2saRef.
  1. a

    Values for individual meteorites (or means for individual meteorites, when available) were compiled for each meteorite group. 2 std. err. = 2 standard error, also called 2 standard deviation mean; this is the 2 s.d. divided by the square root of the number of measurements.

  2. b

    Includes analyses for all types of carbonaceous chondrites reported in b, d, h, and i, some of them not shown in this table.

  3. References: a = Qin et al. (2010a); b = Qin et al. (2010b); c = Shukolyukov and Lugmair (2004); d = Shukolyukov and Lugmair (2006b); e = Shukolyukov and Lugmair (2006a); f = Shukolyukov et al. (2009); g = Trinquier et al. (2007); h = Trinquier et al. (2006); i = Trinquier et al. (2008); j = Yamakawa et al. (2010); k = Yamashita et al. (2005).

Carb. chondritesb0.,d,h,i
E chondrites0.,c,h,i
Angrites  −0.360.01f,h
K/T boundary0.190.060.860.22g
Bulk Earth0.,h,i

This result shows unambiguously that the Cr composition in the El'gygytgyn impact glass G4 is significantly different from terrestrial values, which implies a meteoritic origin for a large fraction of the chromium in this sample (Fig. 4 and values in Table 3). The negative ε54Cr found here is clearly different from all known carbonaceous chondrite values (ε54Cr = 1.05 ± 0.73), which precludes such a type of impactor. The El'gygytgyn ε54Cr = −0.72 ± 0.31 is close to reported values for ureilites (−0.90 ± 0.12) and eucrites (−0.67 ± 0.27), but is also within analytical error of ordinary chondrites (−0.39 ± 0.12) (see Table 3 for data and references). Chromium-53 results are a bit more ambiguous, as ε53Cr in G4 is not statistically different from Earth or meteorite groups except for eucrite, which can therefore be excluded as a possible impactor type.

Figure 4.

ε53Cr and ε54Cr results from this study (diamonds) and for different groups of meteorites (circles): carbonaceous chondrites (CC), enstatite chondrites (EC), SNC meteorites, eucrites, ordinary chondrites (OC), and ureilites. Values for Earth and K/Pg boundary samples from this study (diamonds) and previous work (squares, see references in Table 3) are also reported. Average values and error bars are from Table 3.

Most of the extraterrestrial material reaching Earth is of chondritic composition, and many impact structures have been formed by chondritic projectiles (e.g., Koeberl et al. 2007b). However, based on our results, the El'gygytgyn meteorite was most probably not a carbonaceous chondrite. An ordinary chondrite-type is still possible based solely on ε54Cr; this meteorite type is also favored by Goderis et al. (2013), but only based on data for a few spherules, not based on suevite data from the drill core. The Cr, Co, and Ni abundance ratios in the latter match neither chondritic nor achondritic compositions, as discussed above.

Achondrites are a dominant source of meteorites after chondrites, and ε54Cr for both eucrite (if barely) and ureilite compositions could match the El'gygytgyn data. The PGE abundances and the Os isotopic composition observed in El'gygytgyn impactites by Goderis et al. (2013) argue against a differentiated achondrite source, as these are too depleted in PGEs. Even though Gurov and Koeberl (2004), as well as the earlier work by Val'ter et al. (1982), proposed that the El'gygytgyn meteorite might have been a ureilite, it is difficult to reach a clear conclusion from all the new data—siderophile element abundance data are inconclusive and Os isotopic values not specific to a meteorite type, leaving only the Cr isotope data to provide some constraints. Our results favor a ureilite-type impactor, but do not statistically exclude an ordinary chondrite; however, they positively exclude any other type of meteorite (carbonaceous chondrite, enstatite chondrite, eucrites…).

To account for the ε54Cr value in sample G4 by mixing of bulk Earth material and either ureilite or ordinary chondrite material, most of the Cr in G4 must be of extraterrestrial origin: approximately 40–100% for ureilites, and 80–100% for ordinary chondrites (when considering the whole range of statistical variation for ε54Cr data). Based on Cr concentrations in ureilites between 4000 and 40000 ppm (Warren et al. 2006), a ureilitic source would have to represent 0.02–0.5% of the material in impactite G4. Alternatively, the range of ordinary chondrite (approximately 3500 ppm Cr, Qin et al. 2010b) material that could be mixed in G4 is very narrow at around 0.5% and higher than the 0.05% estimated by Goderis et al. (2013).

Given the good agreement of the Cr isotope data with ureilitic compositions, and the only marginal (even though statistically possible) overlap with ordinary chondrite values, we currently prefer the interpretation of our data as being more consistent with a ureilite impactor. The recent discovery of a small F-type asteroid that subsequently collided with the Earth at Almahata Sitta in Sudan led to the recovery of numerous ureilite fragments (e.g., Jenniskens et al. 2009; Qin et al. 2010a) and indicates that such bodies may also cause impact craters. It is interesting to note that mineralogical studies of Almahata Sitta showed that 70–80% of the approximately 600 fragments recovered are ureilites, whereas the remaining 20–30% are enstatite chondrites and H and L ordinary chondrites (e.g., Shaddad et al. 2010). It is entirely possible that an asteroid with such a diverse composition—which would then result in difficult to interpret geochemical contamination in impact melts and glasses—is not a unique object, as low-velocity collisions in the asteroid belt could lead to some mixing (Gayon-Markt et al. 2012).


Funding for the El'gygytgyn drilling project was provided by the International Continental Scientific Drilling Program (ICDP), the U.S. National Science Foundation (NSF), the German Federal Ministry of Education and Research (BMBF), Alfred Wegener Institute (AWI) and GeoForschungsZentrum Potsdam (GFZ), the Russian Academy of Sciences Far East Branch (RAS FEB), the Russian Foundation for Basic Research (RFBR), and the Austrian Federal Ministry of Science and Research (BMWF). The Russian GLAD 800 drilling system was developed and operated by DOSECC Inc., and the downhole logging was performed by the ICDP-OSG. The current work was supported by the Austrian Science Foundation (FWF), project P21821-N19. We thank E. Gurov (Kiev) for many years of advice and supply of some surface samples from El'gygytgyn, and L. Pittarello and D. Mader (both University of Vienna) for assistance with the petrography and geochemistry of the drill core samples discussed in this article. We are grateful to S. Goderis for discussions, to B. Schmitz and A. Shukolyukov for constructive and helpful reviews, and to W. U. Reimold for editorial comments.

Editorial Handling

Dr. Uwe Reimold