Jiddat al Harasis 422: A ureilite with an extremely high degree of shock melting

Authors


Corresponding author. E-mail: emilie.janots@ujf-grenoble.fr

Abstract–

The Jiddat al Harasis (JaH) 422 ureilite was found in the Sultanate of Oman; it is classified as a ureilitic impact melt breccia. The meteorite consists of rounded polycrystalline olivine clasts (35%), pores (8%), and microcrystalline matrix (57%). Clasts and matrix have oxygen isotopic values and chemical compositions (major and trace elements) characteristic of the ureilite group. The matrix contains olivine (Fo83–90), low-Ca pyroxene (En84–92Wo0–5), augite (En71–56Wo20–31), graphite, diamond, Fe-metal, sulfides, chromite, and felsic glass. Pores are partly filled by secondary Fe-oxihydroxide and desert alteration products. Pores are surrounded by strongly reduced silicates. Clasts consist of fine-grained aggregates of polygonal olivine. These clasts have an approximately 250 μm wide reaction rim, in which olivine composition evolves progressively from the core composition (Fo79–81) to the matrix composition (Fo84–87). Veins crossing the clasts comprise pyroxene, Fe-oxihydroxide, C-phases, and chromite. Clasts contain Ca-, Al-, and Cr-rich glass along olivine grain boundaries (<1 μm wide). We suggest that a significant portion of JaH 422, including olivine and all the pyroxenes, was molten as a result of an impact. In comparison with other impact-melted ureilites, JaH 422 shows the highest melt portion. Based on textural and compositional considerations, clasts and matrix probably originated from the same protolith, with the clasts representing relict olivine that survived, but was recrystallized in the impact melt. During the melt stage, the high availability of FeO and elevated temperatures controlled oxygen fugacity at values high enough to stabilize olivine with Fo∼83–87 and chromite. Along pores, high Mg# compositions of silicates indicate that in a late stage or after melt crystallization FeO became less available and fO2 conditions were controlled by C−CO + CO2.

Introduction

Ureilites are achondritic meteorites that have characteristics both of primitive (chondritic) and of highly differentiated materials (e.g., Mittlefehldt et al. 1998). For example, they have oxygen isotopic ratios that scatter along a trend of slope approximately 1, nearly coincident with the CCAM trend, on the standard 3-oxygen isotope diagram (Clayton et al. 1976; Clayton and Mayeda 1996). And yet their mineralogy and chemical compositions are far from chondritic. Main group ureilites are coarse-grained ultramafic (olivine + pyroxene) rocks with unbrecciated textures. They contain also abundant carbon phases, and accessory metal and sulfide (Goodrich 1992; Mittlefehldt et al. 1998; Goodrich et al. 2004). Interstitial carbon occurs as graphite, diamond, and/or amorphous carbon (Vdovykin 1969; Bischoff et al. 1999; El Goresy et al. 2004). Although ureilite classification is not unanimous (Berkley et al. 1980; Ikeda et al. 2000; Mittlefehldt et al. 2005), the ureilite group has been divided into three subgroups based on pyroxene composition in Goodrich et al. (2004, 2007, 2009): olivine-pigeonite, olivine-orthopyroxene, and augite-bearing. Olivine-pigeonite ureilites are the most abundant group; they have olivine and pigeonite compositions of Fo∼76–87 and Wo∼6–13, respectively, and show increasing pyroxene/pyroxene + olivine (py-ratio) with increasing Fo of olivine (Singletary and Grove 2003; Goodrich et al. 2007). Olivine-orthopyroxene ureilites constitute the most magnesian group, with Fo∼86–92 in olivine, and have the highest py-ratios (Goodrich et al. 2007). In the olivine-pigeonite and olivine-orthopyroxene subgroups, olivine and pyroxene each define an Fe/Mn-Fe/Mg trend of near-constant, chondritic Mn/Mg ratio (Mittlefehldt 1986; Goodrich et al. 1987, 2006). This trend, occurring over such a large range in the Mg#, is interpreted by some workers as a result of redox processes controlled by the abundant carbonaceous material. Other workers proposed that the Fe-Mg compositional spreading may reflect primary heterogeneities within the ureilite parent body (UPB) (Warren 1996; Mittlefehldt et al. 2005; Warren and Huber 2006a; Warren et al. 2006a). The third, small subgroup, the augite-bearing ureilites, diverges from this reduction trend, showing limited igneous fractionation relative to the low-Ca pyroxene groups (Goodrich and Delaney 2000; Goodrich et al. 2004; Downes et al. 2008). Brecciated ureilites contain fragments from all the three main ureilite subgroups, as well as additional indigenous or exotic fragments (e.g., Ikeda et al. 2000; Downes et al. 2008).

The three ureilites Ramlat as Sahmah (RaS) 247 and Jiddat al Harasis (JaH) 422 and Jiddat al Harasis (JaH) 424 were discovered within a perimeter of 50 km during a systematic search for meteorites in the Sultanate of Oman, in 2005 and 2007. RaS 247 consists of 30 vol% olivine, 54% pyroxene (pigeonite) and 16% 0.5–1.8 mm-sized carbon plates (Connolly et al. 2007). JaH 424 is a fine-grained, weakly shocked main group olivine-pigeonite ureilite (Weisberg et al. 2009). While both RaS 247 and JaH 424 belong to the main ureilite group, JaH 422 presents some unique features. It has a reddish heterogeneous matrix containing yellowish clasts, pores, and C-phases including diamond (Fig. 1a). In the present study, the petrography, geochemistry, and oxygen isotopic composition of JaH 422 ureilite are investigated, and the formation of this sample is discussed in the context of the impact history of the UPB. We also investigate bulk chemical compositions of RaS 247 and JaH 424, and use them to help in distinguishing terrestrial contamination from pristine chemical signatures.

Figure 1.

 a) Slab of JaH 422 with millimeter scale. Yellowish-greenish olivine clasts occur in a reddish matrix containing abundant pores and C-minerals (gray). b) to e) are transmitted light optical microscope images: b) Five polycrystalline olivine clasts in a matrix consisting of silicates, oxides, C-minerals, and pores (plane light). c) Euhedral matrix olivine associated with twinned low-low-Ca pyroxene (crossed polars). d) Portion of a large polycrystalline olivine clast (plane light). Some “veins” intersect at triple junctions interpreted as grain borders of an original coarser-grained lithology. e) Detailed view of a vein of Fig. 1d (crossed polars). The vein consists of pyroxene, pores, carbonaceous materials, and Fe-oxyhydroxide and is surrounded by recrystallized olivine of the clast.

Analytical Methods

In JaH 422, point counting was achieved on digital images of the whole slab shown in Fig. 1a (2517 points), by distinguishing three components: yellowish clasts, pores (>50 μm), and matrix (the remaining material). Texture and mineral assemblages were characterized on one single polished thin section. They were first characterized using optical microscopy. Quantitative mineral analyses were determined on one single polished thin section using a Jeol JXA8200 electron microprobe (EMP), at beam conditions of 15 kV, 20 nA, 1 μm spot size, and a ϕ(ρz) matrix correction procedure (Pouchou and Pichoir 1984). Special attention was paid to BSE contrast variations at grain and pore boundaries. Profiles across olivine clasts were measured with a step size of 1 μm and counting times of 900–1900 ms per step. X-ray maps (Ca, Mg, Cr, Fe) were obtained at beam currents of 40 nA, with a dwell time of 60 ms and a step size of around 1 μm. Mineral abundances were roughly estimated from elemental X-ray maps (see protocol below). As a result of small grain size (typically <1–2 μm), oxides, felsic material, and SiO2 have been identified from EDS measurements, as quantitative analyses were compromised by interference from neighboring grains. Representative analyses of olivine, pyroxene, and chromite, from both clast and matrix, and compositions of SiO2-rich glass are compiled in the Supporting Information section. It is noteworthy that chromite size (few micrometer; it appears too large in Fig. 5 as an artifact of X-ray mapping) hampers precise EMP analyses and only few grains gave acceptable analyses (see supporting information). Considering the high analytical uncertainties, all analyses yield comparable compositions in the matrix and clast, indifferently.

Figure 5.

 X-ray element distribution maps in an olivine clast. a) Mg, b) Ca, and c) Cr. Mg and Fe are anticorrelated in the reduction rim. Ca- and Cr-rich glass forms an interconnected network along olivine grain boundaries. Clast polycrystalline texture is the most visible on the Ca-X-ray map (b). Chromite, visible as tiny white patches in (c), crystallizes close to the contact between olivine clast and matrix.

Carbonaceous material was identified by Raman spectrometry. In polished sections of JaH 422, diamond shows enormous relief because of its hardness, clearly distinguishing it from diamond grit used for polishing found in small holes. Diamond was identified by Raman spectrometry in those areas with high relief and independently in carbon grains isolated using HF/HCl. This excludes the possibility that polishing grit was analyzed.

Oxygen isotope analyses were carried out at the Open University using an infrared laser-assisted fluorination system (Miller et al. 1999). Thin rock slices were gently crushed and matrix and clast fragments hand-picked under a binocular. For both matrix and clasts analyzed in this study, two approximately 70 mg chips of each were crushed and homogenized, with analyses undertaken on approximately 2 mg aliquots of these powders. Significant levels of terrestrial contamination were detected in the initial matrix sample and so a second approximately 1 g chip of this material was powdered and an aliquot of this material was pretreated by leaching in ethanolamine thioglycolate (EATG; Cornish and Doyle 1984). This procedure removes iron oxides, hydroxides, and metallic iron, but not silicate-bound iron (Martins et al. 2007). O2 was liberated by heating the powdered samples using an infrared CO2 laser (10.6 μm) in the presence of 210 torr of BrF5. After fluorination, the O2 released was purified by passing it through two cryogenic nitrogen traps and over a bed of heated KBr. O2 was analyzed using a Micromass Prism III dual inlet mass spectrometer. Oxygen isotopic analyses are reported in standard δ notation where δ18O has been calculated as: δ18O = [{(18O/16O)sample/(18O/16O)ref} − 1] × 1000 per mil (‰), and similarly for δ17O using the 17O/16O ratio. Δ17O has been calculated according to the relationship Δ17O = δ17 − 0.52 × δ18O. Published system precision (1σ) (Miller et al. 1999), based on replicate analyses of international (NBS-28 quartz, UWG-2 garnet) and internal standards, is approximately ±0.04‰ for δ17O; ±0.08‰ for δ18O; ±0.02‰ for Δ17O. The quoted precision (1σ) for the meteorite samples is based on two or more replicate analyses.

Chemical analyses of samples powdered in an agate mortar were obtained by inductively coupled plasma–optical emission spectroscopy (ICP-OES, for major elements reported in wt%) and inductively coupled plasma–mass spectrometry (ICP-MS, for trace elements reported as ppm) at Activation Laboratories Ltd., Ancaster, Ontario, Canada. Samples were acid-dissolved after fusion with lithium metaborate/tetraborate. Errors (1σ; Miller et al. 1999) typically are 1–5% for major elements and 10% for trace elements, as estimated from rock standards used for calibration and independently measured standards. Analysis quality was controlled using samples of Allende. Analyses were obtained on aliquots of 0.95 g (JaH 422), 1.41 g (JaH 424), and 1.95 g (RaS 247). Calculation of the matrix composition assumed that matrix and clasts have the same molar volume. The matrix composition was estimated by subtracting 38.2% (clast vol%) of the clast contribution from the whole rock composition (atom%). At a first approximation, we considered olivine clast compositions as homogeneous and we used the average composition of olivine core (atom%) obtained from EMP analyses (Table 1). It is noteworthy that similar calculation using average composition of the clast rim modifies by less than 5% the results on the matrix composition, except for Na2O and TiO2 (10% and 7%, respectively). The matrix composition was then recalculated in oxide weight and renormalized to 100% (ox. wt%). Loss on ignition was neglected from the calculation. As a consequence, higher oxide contents are obtained in clasts and matrix compared with the bulk.

Table 1.   Major elements in ureilites RaS 247, JaH 424, and JaH 422 (whole-rock, clast, and matrix).
Oxide wt%RaS 247aJaH 424aJaH 422aJaH 422 olivinebJaH 422 matrixc
  1. aWhole-rock analyses by ICP-OES.

  2. bElectron microprobe analyses average of core composition.

  3. cComposition recalculated by subtracting olivine clasts (see discussion in text).

  4. dFor simplicity, Fe concentrations assume that all Fe is divalent, although in fact it occurs in metallic, divalent, and trivalent states in RaS 247, JaH 424, and JaH 422.

  5. eLOI = loss on ignition.

  6. n.d. = not determined.

SiO239.6439.1936.0538.9038.99
Al2O30.440.320.190.040.31
FeOd19.1419.1817.3319.4518.28
MnO0.3920.3850.3660.420.38
MgO33.5933.6435.6140.8537.03
CaO1.621.652.220.343.65
Na2O0.080.070.050.020.08
K2O<0.010.05<0.010.010.00
TiO20.0310.0220.0290.020.04
P2O50.080.090.06n.d.n.d.
Cr2O30.510.470.630.650.70
NiO0.240.270.25n.d.n.d.
LOIe3.663.564.2n.d.n.d.
Total101.55100.3097.93100.68100.00
Fe/Mg0.320.320.270.270.28
Fe/Mn48.2149.1946.7546.1747.14
Fe/Cr39.6942.9728.9731.6427.46

Results

Petrography

JaH 422 is a single, rounded, reddish-brown stone with a mass of 61.475 g (Fig. 1a). It has a reddish matrix containing yellowish clasts and round pores (Fig. 1a). Fusion crust was not observed. Point counting of the slab yields the following proportions: 56.5 vol% matrix (including carbonaceous material), 35.1 vol% clasts, and 8.4 vol% porosity. Excluding pores, it yields 61.7 vol% matrix and 38.3 vol% clasts.

The matrix is microcrystalline with grain sizes ranging from 1 to 50 μm. It comprises olivine, low-Ca pyroxene, augite, carbonaceous material, rare spinel, and traces of felsic glass. Olivine is the most abundant mineral. Olivine grains are 5–50 μm in size, euhedral to subhedral. Low-Ca pyroxene and augite are often found in an interstitial position between olivine (Fig. 2a). Low-Ca pyroxene grains rarely exceed 20 μm in size (Fig. 2c), and the largest grains show twinning (Figs. 1c and 1e). Clinopyroxene rarely exceeds a few micrometers in grain size; it commonly rims low-Ca pyroxene (Fig. 2c). Fe-metal and Fe-sulfides are occasionally found as round submicrometric crystals (Fig. 2b). In the majority of cases, however, terrestrial weathering has led to the alteration of the primary Fe-metal and Fe-sulfide to Fe-oxihydroxides (maghemite, magnetite, and goethite). Rare felsic glass occurs interstitially between pyroxenes (Fig. 2b); its composition, characterized using an EDS detector, is rich in Ca, Al, and Si and can locally reach almost pure SiO2 compositions (Supporting Information). C-minerals form aggregates up to 1 mm. Micrometric spinels (approximately 5 μm) are preferentially localized at the interface between clasts.

Figure 2.

 a) to g) are backscattered electron images of JaH 422. a) Euhedral olivine, secondary Fe-oxihydroxides (Fe-OOH) and interstitial orthopyroxene in the matrix. b) Association of felsic glass, SiO2-rich material, and Fe-metal in the interstices between zoned pyroxene grains. c) Matrix pores filled with terrestrial bassanite (Bss) and Fe-OOH. An olivine clast is marked by a black line. d) Interface of pores with matrix marked by strongly reduced Mg-rich orthopyroxene. e) Olivine clast displaying a reaction rim. f) Interior of the largest observed olivine clast (Fig. 1d). Veins in the clasts are filled with an assemblage of orthopyroxene, graphite and diamond (C), and Fe-OOH. See also (h). g) Highly reduced orthopyroxene at the contact with Fe-OOH in a clast vein. h) Sketch of Fig. 1f. The limits of the olivine clast are outlined. Glass occupies fissures bordered by reduction rims (darker gray). These fissures and some of the largest veins (dark gray) are interpreted as the original boundaries of pristine coarse olivine.

In JaH 422, clasts have rounded or angular morphology with a size between 0.3 and 14 mm (Figs. 1a, 1b, and 1d). All clasts are polycrystalline (Fig. 1d). They consist of approximately 50 μm-sized, polygonal olivine grains, frequently showing 120° triple junctions. These olivine crystals show random orientations and are free of shock features. Some veins crosscut the clasts. They are filled by an assemblage of low-Ca pyroxene, Fe-metal and Fe-sulfide, and carbonaceous material. In these veins, Fe-phases occur as tiny blebs or as vein coatings or fillings. Olivine clasts are zoned with a homogeneous core surrounded by a concentric rim. The rim dimensions (approximately 250 μm) are approximately identical in all clasts; this rim completely encompasses the smallest clasts. The contact between clasts and matrix is gradational in backscattered electron (BSE) images or on X-ray maps (Figs. 2e and 5a). Olivine crystals are euhedral at the interface between clasts and matrix. Iron metal and Fe-sulfide occur preferentially in veins cross cutting the clasts. However, olivine clast rims are devoid of metal blebs. Interstitial felsic material is found along the olivine grain boundaries of the polycrystalline clast (Fig. 5b). Micrometric chromite grains are preferentially localized at the interface between clasts and matrix and in olivine clast veins (Fig. 5c). They are usually below 5 μm in size.

In the matrix, roundish pores, up to 1 mm in diameter, are distributed randomly; they do not show preferential orientation or interconnection. They are partly filled with terrestrial alteration products. Pores are commonly surrounded by Fe-oxihydroxides and low-Ca pyroxene.

Mineral Compositions

In the fine-grained matrix, olivine grains (Fo83–87) have Fe-Mg-Cr and Fe-Mg-Mn compositions (Fig. 3a) following the ureilite reduction trend (Goodrich et al. 1987). More reduced compositions (Fo83–92) are found at the contact with C-material, Fe-phases, and pores (Figs. 3a and 3b). Low-Ca pyroxene shows a wide range of compositions (En84–92 and Wo0–5). At the contact with pores, highly reduced low-Ca pyroxene (up to 98% Mg#) has very low Ca contents (Fig. 4). All analyzed high-Ca pyroxenes are augites with a large range in Ca contents (Fs9–12Wo20–31). This is likely the result of unavoidable mixed analyses (augite diameter approximately 1–5 μm). In terms of Fe-Mg-Mn composition (Fig. 3c), low-Ca pyroxene plots above or along the ureilite reduction trend (Goodrich et al. 1987, 2004). Fe-phases in the matrix include Fe-metal, Fe-sulfides, and Fe-oxihydroxides. Ni content in these phases is variable but can reach a few percentage in some of the Fe-metal grains. C-phases were identified as graphite and diamond by Raman spectrometry. Spinel analyses contain high SiO2 and have marginal stoechiometry, because of contamination by adjacent silicates. However, the analyses show a reproducible composition of a Fe- and Al- rich Mg-chromite, Mg0.6–0.7Fe0.4–0.5Cr1.0Al.0.8–0.9O4 (Table S6), and will be referred to as chromite.

Figure 3.

 a) Plot of molar Fe/Cr versus Fe/Mg in olivine. Fe/Mn versus Fe/Mg in olivine (b) and pyroxene (c). Solid lines correspond to the fit of the data (olivine and pigeonite) published in Goodrich et al. (1987, 2004). Fe-OOH = Fe-oxihydroxides.

Figure 4.

 Compositions of pyroxene in JaH 422 clasts and matrix.

Clasts consist of olivine aggregates displaying a homogeneous clast core (∼Fo78–80) and a concentric rim (Fig. 2e). The clast rim is marked by a progressive Mg# increase reaching the composition of matrix olivine (Fig. 5a). Along grain boundaries, interstitial material is enriched in Na, Ca, Al, and Cr (Figs. 5b and 5c and EDS analyses). Quantitative analyses of this material are hampered by the submicrometric size, but qualitative element profiles across the clasts confirm their composition by concomitant Ca, Cr, and Al peaks. In veins within the clasts, pyroxene shows an extreme spread in composition (Mg# = 0.78–0.98). The more reduced compositions are the poorest in Ca and are observed at the contact with the pores (Fig. 2g). One single augite analysis (Fs17Wo22) has a lower Mg# compared with matrix augite. In terms of Fe-Mg-Mn and Fe-Mg-Cr compositions, olivine and pyroxene grains show similar tendencies to those in the matrix (Fig. 3). Olivine and low-Ca pyroxene plot slightly above the ureilitic reduction trends for olivine and pigeonite, respectively (Goodrich et al. 1987, 2004). In the veins, Mg-chromite is similar in composition to that found at the interface between matrix and clasts.

Pores are often filled by an association of secondary Fe-oxihydroxides and desert alteration products, identified by Raman spectrometry as bassanite, 2CaSO4·H2O (Figs. 2c and 2d). Silicates are reduced at the interface between the pores and the matrix (Fig. 2d) and the highest Mg# minerals are found at the contact with pores.

Oxygen Isotopes

Oxygen isotope analyses were undertaken on samples from both clast and matrix material. An initial analysis of the matrix indicated that it had been significantly affected by terrestrial alteration and as a consequence two further replicates were leached in EATG prior to analysis (see the Analytical Methods section). The mean oxygen isotope composition of the two leached matrix replicates was: δ17O = 3.31‰ ± 0.04, δ18O = 8.31‰ ± 0.01; Δ17O = −1.00‰ ± 0.01. The mean isotopic composition of two replicate analyses of untreated clast was: δ17O = 2.87‰ ± 0.02, δ18O = 7.57‰ ± 0.04; Δ17O = −1.07‰ ± 0.00. A single analysis of the untreated matrix material had the following composition: δ17O = 5.23‰; δ18O = 11.84‰; Δ17O = −0.93‰. Figure 6 shows the oxygen isotope composition of both the clast and treated matrix material in relation with other major achondrite lithologies. As can be clearly seen from this diagram, both the clast and matrix samples plot well within the ureilite field and have fairly similar isotopic compositions. This evidence is consistent with both lithologies being indigenous components, rather than being derived from an external source as is sometimes the case in polymict ureilites (Goodrich et al. 2004; Kita et al. 2004; Downes et al. 2008). The untreated matrix sample is significantly displaced to higher δ18O values compared with the leached aliquot (Fig. S1). This relationship is consistent with results for other hot desert finds, which also display significant oxygen isotope shifts to heavier values (Greenwood et al. 2008). As it is unlikely that the EATG treatment completely removes all traces of terrestrial alteration, the closely similar isotopic composition of the clast and treated matrix, with only slightly higher δ18O, suggests that both may be samples of one identical UPB reservoir.

Figure 6.

 Plot of δ18O versus Δ17O showing the oxygen isotope composition of the JaH 422 treated matrix sample and olivine clast in relation to other major achondrite groups (data from Clayton and Mayeda 1996; Franchi et al. 1999; Newton et al. 2000).

Whole-Rock Composition

Major elements concentrations of RaS 247, JaH 424, and JaH 422 (Table 1) are all in the range of known ureilite values (Mittlefehldt et al. 1998). The three meteorites have ultramafic compositions: lithophile incompatible elements are depleted compared with CI chondrite (Fig. 7; Tables 1 and 2). While Al, K, and Na contents of JaH 422 are similar to those of ureilites recently reviewed by Warren and Huber (2006b), Sr, Ba, and U are higher (12–186 × CI; Table 2). In comparison with RaS 247 and JaH 424, JaH 422 shows slightly lower Si, Fe, and Mn concentrations and Mg and Cr enrichment. All three ureilites analyzed here are characterized by a CI normalized rare earth element (REE) pattern displaying a progressive decrease of the light REE and a flattening of the heavy REE (Fig. 8). REE in JaH 422 are in the range of olivine-pigeonite ureilites. The calculated matrix composition is also typical of ureilite composition, except for the high CaO content (see procedure details in the Analytical Methods section).

Figure 7.

 CI- and Mg-normalized lithophile element abundances of JaH 422, of the two ureilites RaS 247 and JaH 424 found in same area, of olivine-pigeonite ureilites (Allan Hills [ALH] A77257, Haverö, Kenna, Novo Urei), and of an augite-bearing ureilite (LEW 88774). Apart from Omani ureilites, data are from Mittlefehldt et al. (1998).

Table 2.   Whole-rock trace elements in ureilites RaS 247, JaH 424, and JaH 422.
Element (ppm)Detection limit (ppm)RaS 247JaH 424JaH 422
Sc1987
V5857768
Sr21131352754
Ba3309742
Cr20349032304330
Zn3025090350
Rb122<1
La0.050.370.640.14
Ce0.050.581.170.23
Pr0.010.090.120.03
Nd0.050.310.330.07
Sm0.010.060.06<0.01
Eu0.0050.0060.009<0.005
Gd0.010.090.09<0.01
Tb0.010.020.01<0.01
Dy0.010.090.090.03
Ho0.010.020.02<0.01
Er0.010.080.080.08
Tm0.0050.0130.011<0.005
Yb0.010.090.090.04
Lu0.0020.0130.0190.007
U0.010.490.330.26
Figure 8.

 CI-normalized REE pattern of JaH 422 and ureilites RaS 247 and JaH 424 found in same area. For comparison, data of olivine-pigeonite ureilites (ALHA77257, Haverö, Kenna, Novo Urei) and a characteristic pattern of an augite-bearing ureilite (LEW 887774) are added (Mittlefehldt et al. 1998).

Discussion

JaH 422: A Singular Ureilite

JaH 422 has oxygen isotopic composition, whole-rock chemical composition, and mineralogy consistent with those of known ureilite meteorites. Its chondrite-normalized enrichment of Sr, Ba, and U, as compared with documented ureilites (Mittlefehldt et al. 1998) can be explained by the presence of terrestrial alteration phases like the Ca-sulfate bassanite, and sorption on Fe-oxihydroxide (Al-Kathiri et al. 2005). Ca, Sr, Ba, and U are commonly high in hot desert meteorites and indicate significant desert alteration (e.g., Stelzner et al. 1999); they are similarly enriched in the two other Omani ureilites, RaS 247 and JaH 424, found in the same region. However, in several aspects of texture and mineral composition, JaH 422 is unusual compared with either main group or polymict ureilites. It displays matrix and clasts with a microcrystalline texture, significant porosity, and a spread in silicate compositions. Some of these features have already been reported in highly shocked meteorites (Saito and Takeda 1990; Warren and Rubin 2006, 2010). They were especially observed in ureilitic lithologies of the recent description of some fragments of the Almahata Sitta fall (Jenniskens et al. 2009; Zolensky et al. 2009; Warren and Rubin 2010) and in the Larkman Nunatak (LAR) 04315 ureilite (Warren and Rubin 2010). However, it appears that Almahata Sitta is a heterogeneous breccia containing many different ureilitic and chondritic lithologies (Bischoff et al. 2010).

Extreme Shock Conditions Recorded in Ureilite JaH 422

JaH 422 contains highly shocked clasts in a matrix with igneous texture; we propose that it is an impact melt breccia. The matrix displays several features indicating that melting took place: pores, finely recrystallized magmatic textures, and relict areas of felsic melt (Figs. 1 and 2). The pores have rounded morphology with walls composed of the most reduced silicates (Figs. 2c and 2d). Reduced silicate compositions result from reduction reactions during melting (see below). Hence, association of pores with the most reduced silicates precludes the possibility that the porosity was created by desert alteration.

The coarse-grained texture and zoning of the polycrystalline olivine clasts indicate that they are highly shocked and recrystallized, but did not experience impact melting. They are analogous to the fine-grained olivine aggregates observed in numerous highly shocked ureilites (Berkley et al. 1980; Saito and Takeda 1990; Goodrich 1992; Rubin 2006). Although some of these olivine aggregates were first identified as mosaicized (Goodrich 1992), they have been reinterpreted by Rubin (2006) as the result of annealing after mosaicization. In JaH 422, the grain size (up to 50 μm) and the unstrained, equant, and subeuhedral texture of the olivine are consistent with solid-state recrystallization rather than mosaicism. Within the clasts, olivine grains have a selvage of Ca-, Al-, Cr-, and Si-interstitial material. Similar interstitial material was also observed between coarse olivine grains (Ikeda 1999) or along highly shocked polycrystalline olivines in Lewis Cliff (LEW) 86216 (Saito and Takeda 1990) and LAR 04315 (Warren and Rubin 2006, 2010). This interstitial glass is likely provided by recrystallization of the pristine olivine (Warren and Huber 2006b; Warren et al. 2006b). Based on the shock nomenclature in chondrites (Stöffler et al. 1991), the fine-grained aggregate texture is ascribed to shock stage S6 (Bischoff et al. 1999) or likewise to annealing of mosaic olivine (recrystallization of stage S5 rocks, Rubin 2006).

From the textural relationships, the JaH 422 matrix displays the following crystallization sequence: olivine grains crystallize (I) and are armored by low-Ca pyroxene (II) and Ca-rich pyroxene (III). The remaining felsic melt is quenched (IV). The felsic matrix material resembles felsic interstitial material viewed as an impact melting product (Takeda 1987, 1989; Ogata et al. 1991; Ikeda 1999; Rubin 2006) or as a postdecompression reduction product (Warren and Huber 2006b; Warren and Rubin 2006) in other ureilites. Rubin (2006) proposed that it was injected during the impact. Similarly, Warren et al. (2006b) proposed that the feldspar and feldspathic melt found in Fe,S-rich vein of Elephant Moraine (EET) 96001 were emplaced into the monomict ureilite at an advanced stage in a chaotic disruption and partial reassembly process. JaH 422 matrix contains abundant olivine (approximately 50 vol% of the matrix) and pyroxene. The euhedral texture of matrix olivine suggests that it is not inherited and that it crystallized from the ultramafic impact melt. Coexistence of olivine and SiO2-rich melt can be explained by disequilibrium fractional crystallization. Assuming olivine preservation, without reequilibration after crystallization, the SiO2 concentration increases progressively in the residual melt up to the eutectic, where enstatite and quartz coexist (see fig. 5.8 in Best 2003). Enrichment in SiO2 is enhanced by concomitant reduction reactions (Ikeda 1999; Warren and Rubin 2006). A similar scenario involving disequilibrium crystallization in an impact melt was also suggested by Ikeda (1999) to explain the crystallization of interstitial silicate melt in Asuka-881931. Similarly to the observation in Asuka-881931, in JaH 422 the interstitial glass in contact with matrix pyroxene is distinct from felsic material in contact with olivine clasts in JaH 422. The most Si-rich composition is found in the matrix and a SiO2-phase is only found between matrix pyroxenes (Fig. 2b). Based on experimental results on ordinary chondrite (Stöffler et al. 1991), extreme shock conditions (S6; P > 90 GPa) must have prevailed to melt an olivine-pyroxene rock and to produce an impact melt breccia with the mineral assemblages and textures of JaH 422.

The origin of diamond in ureilites is still unclear. Diamond may grow from vapor (Fukunaga et al. 1987) or by high shock pressure during impacts (Bischoff et al. 1999; Grund and Bischoff 1999; Hezel et al. 2008). Although many workers favor the second possibility for ureilites, it remains unclear whether the impact occurred during or after the UPB differentiation. In JaH 422, diamonds are aggregated in pockets in the matrix, with no obvious evidence of pristine relics. Pressures inferred for impact melting in JaH 422, based on the nomenclature in Stöffler et al. (1991), are compatible with diamond formation at shock pressure >60 GPa (Decarli and Jamieson 1961). However, it cannot be ruled out completely that the diamonds are inherited from earlier stages (before or during UPB differentiation).

JaH 422 Protolith

One fundamental question is whether the clasts and the matrix in JaH 422 derived from a single lithology. Usually, ureilite breccias are polymict assemblages of minerals or lithic fragments (Prinz et al. 1983; Bischoff and Stöffler 1992; Mittlefehldt et al. 1998; Goodrich et al. 2004; Bischoff et al. 2006). Most of the clasts are comparable with the lithologies found in monomict ureilites (Ikeda et al. 2000; Downes et al. 2008). Only one case of a dimict breccia has been reported (Smith et al. 2000; Fioretti and Goodrich 2001). In JaH 422, clasts are exclusively composed of fine-grained olivine aggregates. No coarse-grained pyroxenes were found. One possibility is that JaH 422 is a dimict breccia with shocked olivine clasts enclosed in an impact melt. To have only one clast type in one impact melt breccia is, however, rather unrealistic. It is thus assumed that other mineral or lithic fragments were originally present but were dissolved in the ultramafic impact melt. A second possibility is that JaH 422 is an incompletely molten rock, where the olivine “clasts” are remnants of the original protolith. Although matrix and clast appear different at a first glance, their distinction may be inherited from impact. This assumption is consistent with the similar oxygen isotopic compositions between the clasts and matrix, where slightly higher δ18O value of the matrix is presumably attributed to terrestrial alteration (Fig. S1). The whole-rock composition of JaH 422 falls in the range of the most ferroan ureilite subgroup (Fe-rich group 1 in Mittlefehldt et al. 2005). This suggests that the olivine-pigeonite ureilite group is a potential candidate as JaH 422 precursor. Even with the olivine clasts subtracted, the calculated matrix composition remains in the most ferrous ureilite subgroup (Table 1). Low Mg# in olivine clast cores (∼Fo78–80) is also consistent with an olivine-pigeonite lithology as precursor. In this second scenario, the lack of coarse pyroxene relicts may be explained by the small initial py ratio of olivine-pigeonite protolith. Based on matrix and clast abundances, at least 61.7 vol% of the JaH 422 precursor was molten during impact. JaH 422 is certainly the ureilite with the highest degree of melting observed to date.

JaH 422 Conditions During and After Impact

The UPB has a complex thermal history, which is, at least partly, controlled by collisions (review in Rubin 2006). Among possible scenarios for the UPB differentiation, several authors even invoke an impact melt process (e.g., Takeda 1987; Rubin 1988). The complex thermal and shock history of the UPB has been divided by Rubin (2006) in four periods (1) formation; (2) initial shock; (3) postshock annealing and finally (4) postannealing shock. Texture and composition of coarse-grained olivine and pyroxene attest to equilibration at high temperature in the earlier stage 1. There is a consensus that the original ureilite parent body experienced a sudden drop of temperature (Berkley et al. 1976; Mori and Takeda 1983; Miyamoto et al. 1985), probably caused by impact (Berkley et al. 1980; Warren and Kallemeyn 1989; Goodrich et al. 2004), attributed to stage 2 by Rubin (2006). Extreme shock conditions are recorded within many specimens from the ureilite main group (Saito and Takeda 1990; Goodrich 1992; Rubin 2006) and all ureilites have undergone minimum S2 shock stage or beyond. JaH 422 records extreme shock conditions, which postdate UPB formation and are attributed to the Rubin’s period 2 (Rubin 2006). Following this impact event, it is proposed that polymict ureilites have reaggregated by lithification in a regolith environment at the periphery of the original UPB, or on secondary offspring bodies (see review in Bischoff et al. 2006). It is still unclear as to whether the UPB remained largely intact or was totally disrupted by this impact event (see discussion in Goodrich et al. 2004). Modifications registered in JaH 422 can be used to better characterize this major impact event and implications for the UPB history.

To explain fine-grained polycrystalline aggregates, Rubin (2006) invoked two distinct collisions, the first for shock (olivine mosaicism) and the second for annealing of the deformation features. In JaH 422, one collision seems sufficient to produce the fine-grained aggregates. The inferred shock conditions (S6) are high enough to provoke olivine recrystallization. At the contact with the matrix, clasts have a corroded shape indicating that dissolution took place, but olivine grains are euhedral. These observations suggest that clasts first dissolved in the matrix and then recrystallized along with the matrix. Experimental work shows that approximately 10 μm-sized polygonal, recrystallized olivines form during postshock annealing, if shock pressures are sufficiently high (Reimold and Stöffler 1978). However, it remains unclear whether approximately 50 μm-sized polygonal olivine grains, as observed in JaH 422, can be produced by fast cooling following impact melt, or whether some tempering in an impact melt pool is required. In the main ureilite group, a sudden drop in both temperature and pressure is recorded in the chemically reduced rims of olivine and the compositional variation of pyroxenes (Berkley et al. 1976; Mori and Takeda 1983; Miyamoto et al. 1985). In JaH 422, disequilibrium textures are also in agreement with a rapid cooling following the impact. Rapid cooling can also account for low-Ca pyroxene twinning (Figs. 1c and 1e), because of protoenstatite to clinoenstatite transformation (Yasuda et al. 1983). In some aspects, the reduced rims in JaH 422 differ from that of the main group ureilites. Usually, reduced rims of olivine contain abundant tiny Fe-metal blebs formed during smelting reactions. However, these were not observed in JaH 422 (Figs. 1b, 2e, and 5e). Furthermore, the rim size (approximately 250 μm) is rather large compared with most commonly reported reduction rims of 10 to 100 μm (Mittlefehldt et al. 1998). Finally, the reduced rims in JaH 422 only reached Fo∼83–86 and are less reduced than most of the reduced rims in the main ureilite group (see discussion below). In JaH 422, these divergences can be explained if the FeO-rich impact melt (derived from a protolith rich in olivine with Fo80) dominated the mass balance and precluded very low oxygen fugacity (fO2) conditions during the melt stage. However, it is unclear whether Fe-Mg diffusion in polycrystalline clasts in the presence of melt is fast enough to support the well-accepted cooling history displayed by the main group ureilites (3–20 °C h−1, Mori and Takeda 1983; Miyamoto et al. 1985).

In JaH 422, Mg# in silicates indicates more reduced compositions in the matrix compared with the clasts. This reduction of the melt is seen as the result of smelting reactions (Walker and Grove 1993; Goodrich et al. 2007) affecting the bulk melt. The pores are viewed as remnants of gas bubbles, presumably formed by oxidation of carbon phases when the matrix was still partly molten. One simplified reaction can be written as:

image(1)

Smelting reactions are strongly influenced by pressure. Experiments (Walker and Grove 1993) and calculations (Goodrich et al. 2007) showed that pressures of approximately 100 bars are required to maintain equilibrium between graphite and olivine with Fo∼78, similar to those observed in olivine cores in the clasts. In the JaH 422 matrix, most olivine grains have a restricted composition of Fo83–86. However, highly reduced assemblages containing Opx (Mg# up to 98) are found locally in narrow, μm-sized zones around C-material, altered Fe-metal, or pores. In these zones, Mg# compositions are similar to the most reduced rims of silicates of the main group ureilites, which are interpreted as incomplete (solid-state) reequilibration during excavation and rapid cooling. Within the JaH 422 matrix, the observed coexistence of two Mg# can be explained by a switch of fO2 from FeO-controlled conditions (corresponding to buffering by olivine with ∼Fo80) to much more reducing conditions following completion of olivine and pyroxene crystallization, when FeO availability was limited by solid state diffusion and fO2 conditions close to C−CO + CO2 must have prevailed. We propose that, after melt solidification, smelting reactions were confined around pore space and locally led to highly reduced compositions. This scenario assumes that C phases were not in equilibrium with the silicate melt and did not control the oxygen fugacity at this stage. Possible factors limiting C phase oxidation are slow kinetics (abundant diamond) and the possible formation of carbide coatings on C phases. Assuming equilibrium of the impact melt with C phases, a matrix olivine composition of Fo83–86 would only be compatible with a pressure of approximately 70 bars (fig. 4 in Walker and Grove 1993). Such a pressure would mean deep (several 10’s of km) burial, inconsistent with the low-pressure smelting observed in the silicates around pores, and with other evidence of relatively rapid cooling.

In JaH 422, impact melt stabilized trivalent Cr, at least in chromite. In the main group ureilites, Cr is normally found in the divalent form (Warren and Huber 2006b; Warren et al. 2006b). Rare chromite occurrences were documented in peculiar ureilite specimens, where higher oxygen fugacity conditions are inferred for the equilibration (Fo < 77: Prinz et al. 1994; Goodrich 2001; Sikirdji and Warren 2001). In JaH 422, chromite crystallizes at fO2 conditions in the impact melt corresponding to Fo83–86. Assuming an average XFa of 0.15 for JaH 422 matrix olivine, −logfO2 change as a function of temperature can be estimated (e.g., Williams 1972): it evolves from 13.2 at 1250 °C to 9.7 at 1650 °C, close to melting temperature of pure olivine with Fo80. Thermal increase, required to melt JaH 422, likely changed the Cr valence as well as the Cr solubility (Hanson and Jones 1998). While Cr2+ predominates for redox conditions <IW−1, Cr3+ and Cr2+ coexist in equivalent concentrations at −logfO2 approximately 10. Furthermore, Cr2+ preferentially partitions into the melt with increasing temperature. In JaH 422, most of the chromium was likely fixed in the divalent form in silicates before impact. There are evidences that Cr is redistributed during olivine recrystallization: chromium is enriched in interstitial material of the polycrystalline clasts and chromite crystallizes locally (Fig. 5). This partitioning likely accounts for the slight deviation of the Fe/Cr versus Fe/Mg in comparison with the reduction trend (Fig. 3). Hence, it is likely that thermal increase as a result of impact, combined with high FeO availability in melt, favored chromite crystallization in JaH 422. It is noteworthy that chromite also occurs in association with impact melt in Asuka-881931 (Ikeda 1999) and Northwest Africa (NWA) 766 (Sikirdji and Warren 2001).

Conclusions and Implications

JaH 422 is a ureilite that shows a higher degree of shock melting than previously observed ureilite impact melt breccias. The matrix is similar to impact melt found in main group ureilites, but with much higher melt proportion and abundant olivine. It is interpreted as the result of impact melting of ultramafic ureilite material followed by disequilibrium crystallization within the impact melt. We cannot rule out that polycrystalline olivine clasts were breccias trapped in the impact melt, but we favor the hypothesis that polycrystalline olivine clasts are unmolten relics of JaH 422 precursor material. Shock conditions (S6) were high enough to cause recrystallization of olivine clasts, without requiring annealing during a subsequent collision. In the matrix, the Mg# in silicates indicates that oxygen fugacity during melt crystallization was not controlled by C phases but by FeO in melt/silicates, consistent with the occurrence of chromite. After crystallization of ferromagnesian minerals, FeO availability was much lower, leading to lower fO2 conditions eventually controlled by C−CO + CO2. This led to the formation of very reduced silicate rims around pores.

Acknowledgments

Acknowledgments— This work was partially financed through Swiss National Foundation research grants 21-64929.01, 21-26579.89, 200021-103479/1 and 200020-107681. Dr. Salim Al Busaidi, Ministry of Commerce and Industry, Muscat, is thanked for his support during the project. We appreciated the suggestions of the anonymous reviewer, the constructive comments of Barbara Cohen and we are sincerely grateful for the enlightened and critical reviews from P. Warren and the associate editor, C. Goodrich. We also thank G. Van den Bleeken, I. Mercolli, M. Engi, and K. Bermingham for the fruitful discussions.

Editorial Handling— Dr. Cyrena Goodrich

Ancillary