H/L chondrite LaPaz Icefield 031047 – A feather of Icarus?

INTRODUCTION

In 2003, the Antarctic Search for Meteorites (ANS-MET) recovered LaPaz Icefield (LAP) 031047, a 3.5 Â 2.0 Â 1.5 cm and 16.47 g meteorite ( Fig. 1a and b). Initially, a fine-grained, uniform texture that includes numerous voids and rounded metal-sulfide particles was interpreted to constitute a partly to completely impactmelted L chondrite (Satterwhite and Righter, 2006;Connolly et al., 2007). Such bulk impact melted rocks result from hypervelocity collisions, which induce shock pressures >75 GPa and minimum post-shock temperature increases of $1500°C (Stö ffler et al., 1991;Keil et al., 1997). As recorders of major collisional events, impact melt rocks chronicle the evolution of our Solar System and manifest evidence for the dynamic interaction of celestial bodies. Intriguing examples are L chondrite impact melts: petrologic examinations coupled with Ar-Ar radioisotopic dating of these rocks facilitated the recon-0016-7037/$ -see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10. 1016/j.gca.2011.07.037 struction of the L chondrite parent asteroid's collisional history (e.g., Taylor et al., 1979;Bogard, 1995;Bogard et al., 1995;Kring et al., 1996;Benedix et al., 2008;Weirich et al., 2010).

Interestingly, initial olivine and low-Ca pyroxene compositions reported for LAP 031047 (Satterwhite and Righter, 2006) indicate an intermediate composition with olivine (Fa 23 ) typical for L chondrites, while its low-Ca pyroxene (Fs 17 ) falls within the range of H chondrites (Brearley and Jones, 1998). A potentially extraordinary nature of LAP 031047 is also hinted at by its high porosity. For example, Wilkison et al. (2003), Consolmagno et al. (2008), and Sasso et al. (2009) inferred regolith processes or primordial physical states of asteroids from unusually high porosities of certain ordinary chondrite specimens.

We studied the composition, petrology, physical properties, and K-Ar system of LAP 031047 in order to understand its origin and implications for putative parent asteroids with ordinary chondrite-like composition and matching physical properties.

SAMPLES AND METHODS

Weathering grade A was assigned to LAP 031047, which translates to "minor rustiness; rust haloes on metal particles and rust stains along fractures are minor". Its fracturing grade A stands for "minor cracks; few or no cracks are conspicuous to the naked eye and no cracks penetrate the entire specimen" (Satterwhite and Righter, 2006). The broken meteorite was inspected macroscopically in the curatorial facilities at NASA-Johnson Space Center in Houston after some sample material had been distributed ( Fig. 1a and b).

Physical properties

The porosity of a 993 mg, 0.370 cm 3 chip of LAP 031047 was measured by a combination of helium ideal gas pycnometry (for grain density) and X-ray microtomography (for bulk density), using the methods outlined in Sasso et al. (2009).

The porosity of the sample was also determined by synchrotron X-ray microtomography (lCT) with data collected at a resolution of 15.9 lm/voxel (a voxel is a 3D pixel or volume element). During the digital data extraction with the Interactive Data Language routine BLOB3D (Ketcham, 2005), some gentle smoothing of the data is necessary. We used a slight median smoothing and rejected components of interest that were smaller than five contiguous voxels to eliminate potentially erroneous data. These steps reduced the minimum observable volume to 29 lm/voxel.

The 0.993 g sample chip of LAP 031047 had a bulk volume of 0.370 ± 0.008 cm 3 and a bulk density of 2.69 ± 0.06 g/cm 3 . This value is much lower than typical or-dinary chondrite values of 3.22-3.42 g/cm 3 (Consolmagno et al., 2008), but reasonable given the very porous nature of the sample.

Ideal gas pycnometry yielded a grain volume of 0.27 ± 0.02 cm 3 , a grain density of 3.67 ± 0.23 g/cm 3 , and a porosity value of 27.1 ± 5.1 vol%. This agrees with the 26.7 vol% voids in thin section LAP 031047,4 determined by point-counting. Synchrotron X-ray microtomography resolved a porosity of 11.6%. The minimum visible pore size for synchrotron X-ray microtomography is 2.41 Â 10 À5 mm 3 , indicating that 57% of the pore space detectable by He pycnometry must reside in pores smaller than this cutoff.

Thin section analyses

Two petrographic thin sections, LAP 031047,4 and LAP 031047,9, each capturing sample areas of 1.6 cm 2 (Fig. 2), were made available for study by the Meteorite Working Group. Both thin sections were examined under an optical microscope, a scanning electron microscope and an electron microprobe.

Point-counting under reflected light of 2127 points on thin section LAP 031047,4 was conducted at 500Â magnification and a step-size of 100 lm. The discriminated components are silicates, spinel, Fe-Ni metal and troilite, unidentified alteration features, and voids. Relative sizes of troilite and Fe-Ni metal were recorded in arbitrary bins. Because the nature of sub-lm particles could not be discriminated, modal proportions of metal and troilite in that size bin are reported as combined metal and troilite (Fig. 3).

Geochemical data were obtained from electron microprobe analysis at the NASA-Johnson Space Center in Houston, using a Cameca SX-100 with five wavelengthdispersive spectrometers. Peak positions and intensities of individual elements were calibrated and monitored with well-characterized standards from the NASA-Johnson Space Center standard collection. Silicate minerals were analyzed using an accelerating voltage of 15 keV, a beam current of 15 nA, and a beam diameter of 1 lm. To reduce volatilization, counting times for the measurements were 10 s for Na and K, but 20 s for Si, Fe, Mg, Ca, Cr, Al, Ti, Mn, Ni and P. For analyses of Fe-Ni metal and sulfide, an accelerating voltage of 15 keV and a beam current of 20 nA was used to determine the abundances of Fe, Ni, Co, Cr, Mn, Cu, Ti, Si, V, P, Ca, Mg, and S. Cobalt concentration values were corrected for the interference of the Fe-K b X-ray peak with the Co-K a X-ray peak.

Whole-rock trace element geochemistry

Following the standard HF/HNO 3 and HClO 4 microwave digestion methods outlined in Friedrich et al. (2003), whole rock analyses were performed on a Thermo Electron X Series II inductively coupled plasma mass spectrometer (ICPMS). This yielded abundances of 45 trace elements in three sample aliquots with masses of 121.5, 120.8, and 123.8 mg, each of which were analyzed in duplicate.

Oxygen isotopes

Oxygen isotope ratios of 2 to 3 mg aliquots of bulk meteorite were measured at the Geophysical Laboratory, Carnegie Institution of Washington using the procedure of Rumble and Hoering (1994). Reported values are given as d 18 O, d 17 O, and D 17 O. The d values describe the per mil (&) deviation from the international standard, V-SMOW:

where X ¼ 18 or 17:

The deviation from the terrestrial fractionation line is described as:

Prior to oxygen isotope analysis, crushed bulk rock was ultra-sonicated in dilute hydrochloric acid and magnetic material was removed with a hand magnet. Samples were loaded in a Sharp-type reaction chamber (Sharp, 1990). Successive, repeated blanks were carried out by exposing weighed sample aliquots to 30 torr pressure of BrF 5 vapor for 12 h until there was <150 lm non-condensable gas pressure remaining after a blank run. Oxygen was released quantitatively from bulk rock by heating samples individually with a CO 2 laser in the presence of 30 torr pressure of BrF 5 vapor. Gore Mountain garnet (USNM 107144) was analyzed twice during each analytical session to standardize analyses. Repeated analysis of Gore Mountain garnet, USNM 107144, in comparison to UWG-2 (Valley et al., 1995), gives a d 18 O value of 5.91 ± 0.14 per mil and d 17 O of 3.11 ± 0.07 per mil (V-SMOW).

Comparison with data for equilibrated ordinary chondrites from Clayton et al. (1991) indicates that the average Table 2 Oxygen isotope data for three sample splits of LAP 031047 and comparison of the mean of these data with the range of compositions of ordinary chondrites from Clayton et al. (1991) (Table 2). Compared to other ordinary chondrites, two of the three data points for LAP 031047 plot in-between the fields for H and L chondrites, while one data point falls within the field for H chondrites (Fig. 4).

Ar-Ar radioisotopic dating

Three samples of LAP 031047,5 were irradiated at the University of Michigan Ford Reactor, along with several other LaPaz Icefield samples (Swindle et al., 2009), and analyzed using a VG5400 noble gas mass spectrometer at the Lunar and Planetary Laboratory, University of Arizona. Samples of hornblende MMhb-1, K 2 SO 4 , and CaF 2 were spaced throughout the radiation package to monitor Ar isotope production rates, and allowed us to account for any irregularity of the neutron flux. Analyses of MMhb-1 yielded a J factor of 2.144 Â 10 À3 ± 0.004 Â 10 À3 . After analysis, data were corrected for blanks, decay, and reactor interference. As in Swindle et al. (2009), we corrected for spallation-produced isotopes by finding the minimum 36 Ar/ 37 Ar ratio of all steps and assuming all 36 Ar in that step was spallogenic. Correction for terrestrial atmospheric contamination was attained by assuming a terrestrial ratio of 40 Ar/ 36 Ar for all steps up to and including the one with the lowest 36 Ar/ 39 Ar ratio, amounting to four (A and C) or six (B) steps. We note that for all three samples, the first two extractions had a 40 Ar/ 36 Ar ratio within 2r of the terrestrial value, giving us confidence that such a correction was the appropriate one, rather than assuming solar wind or chondritic trapped Ar. Also, the apparent age was below 200 Ma even before correction for terrestrial Ar in every case. Preliminary Ar-Ar data has appeared in abstract form (Weirich et al., 2008).

All three sample chips of LAP 031047 suggest ages <200 Ma, although none give a plateau (Fig. 14). In each case, the sum of the Ar in all extractions leads to calculated degassing ages of 116 ± 35, 90 ± 45, and 101 ± 5 Ma, so we take 100 ± 55 Ma as our best estimate of the age. Several extraction steps yield apparent ages between 200 and 300 Ma, but the only apparent ages >300 Ma are in high-temperature (>1300°C) extractions releasing the final <1% of the 39 Ar (Fig. 14), where small gas amounts and high temperatures make background corrections uncertain. This suggests that at some time in the recent past, the meteorite underwent complete, or nearly complete, degassing due to very high temperatures that reset its Ar-Ar reservoir.

As well as having an extremely young 40 Ar-39 Ar degassing age, LAP 031047 also appears to have a very young cosmic ray exposure age. We estimated the exposure age two different ways. First, we assumed that all of the 36 Ar was cosmogenic, and found exposure ages of 0.3-0.5 Ma. However, the 40 Ar/ 36 Ar ratios in the early extraction steps from each sample were consistent with terrestrial atmosphere, so the 36 Ar in those steps (15-40% of the total) is almost certainly not cosmogenic, so these are upper limits. Second, to verify that the low calculated exposure age was not a result of a calibration error, we compared the LAP 031047 samples to eight H chondrites studied by Swindle et al. (2009). If we assume (as we did for the 40 Ar-39 Ar data reduction) that the step with the minimum observed 36 Ar/ 37 Ar ratio has no 36 Ar from any source other than spallation, the 36 Ar/ 37 Ar ratio should be proportional to the exposure age, and inversely proportional to the production rate of 37 Ar from Ca in the reactor (which in turn should be proportional to the J factor of the irradiation). Hence, the product of the minimum 36 Ar/ 37 Ar ratio in a sample and the J factor should be roughly proportional to exposure age. Based on this, LAP 031047 appears to have an exposure age a factor of 20-50 lower than any of the H chondrites studied -since many H chondrites have exposure ages of 6-8 Ma (Graf and Marti, 1995), if even one of those eight has such an age, this again suggests an exposure age of LAP 031047 of 0.1-0.5 Ma. There is no evidence for a solar wind contribution to the Ar. However, this, like a determination of the true exposure age, would be much more easily determined by analysis of other noble gases in a sample that had not been irradiated by neutrons. LAP 031047 is an ordinary chondrite with an intermediate H/L chemical composition, very high porosity, and a young degassing age. In order to explore its origin and formational history, we discuss its petrologic characteristics and compare it with other ordinary chondrites.

RESULTS

Tomography

The total amount of Fe-metal visible within our tomography volumes amounts to only 2.85%. This is the porosity corrected volume. The total amount of FeS visible within tomography only amounts to 1.91%. Both of these values are considerably less than the amounts found with traditional point counting techniques. The discrepancy can in part be accounted for by the resolution of the lCT technique. After digital processing, only a volume of about 29 lm 3 can be resolved but point counting determined a full two thirds of the sulfide particles are 631 lm at their maximum length. As for the metal, our low lCT value is more difficult to account for. Potentially, our small sample may be inhomogeneous with respect to larger rocks.

Metal grain size number densities (the number of metal particles smaller in size than any individual given particle versus the particle size) may help classify LAP 031047. Prior data of this type (Friedrich et al., 2008;Sasso et al., 2009) showed that the slopes are 0.53 ± 0.05 for H (4 analyses), 0.77 ± 0.08 for L (33 analyses), and 1.01 ± 0.09 for LL (2 analyses). For LAP 031047, the metal grain size number density versus particle volume slope is 0.66. This places Rubin and Jones (2003) and assuming the same proportion of troilite versus metal in the particles <1 lm. c wt%, recalculated as in "b" but assuming all points counted in particles <1 lm are troilite; normative average compositions of H, L and LL chondrites after McSween et al. (1991). d Excludes apatite and ilmenite.

the number density of particles within LAP 031047 between the H and L chemical groups.

Modal and normative composition

The modal composition of LAP 031047,4 is 61.1 vol% silicate and apatite fragments, 26.7 vol% voids, 0.4 vol% chromian spinel, 5.7 vol% Fe-Ni metal, 1.8 vol% troilite, and an additional 4.2 vol% of metal and sulfide in particles too small to distinguish (<1 lm). If we assume the same proportion of metal to sulfide in these small particles, then the modal composition of the meteorite is 8.9 vol% Fe-Ni metal, 2.9 vol% troilite, 61.1 vol% silicate and apatite fragments, 26.7 vol% voids, and 0.4 vol% chromian spinel.

The recalculated abundances of metal and troilite in LAP 031047 (Table 1) suggest an excess in metal if particles smaller than 1 lm are assumed to exhibit the same metal/ troilite ratio as the particles larger than 1 lm. If an excess of troilite is assumed in this smallest particle-size class, which is suggested by detailed microscopic investigation, the modal abundance of metal and troilite is roughly in accordance with the normative abundances of these components in H chondrites (McSween et al., 1991). The latter assumption is also supported by the metal-sulfide size distri-bution histogram (Fig. 3), which shows that troilite occurs preferentially in smaller particles than metal. The abundance of chromite in LAP 031047 is typical for ordinary chondrites and the recalculated abundances of silicate and siderophile components are grossly similar to H chondrites but dissimilar to L and LL chondrites.

Bulk trace element abundances

All elemental analysis results reported (Tables 3 and 4; Fig. 5a and b) have sample replicate errors of 612% RSD except were 1r errors are noted. The upper bound of the Zn value is due to the extreme heterogeneity of Zn in the aliquots analyzed. For that value we have reported the 2r upper limit, since the ±1r limit would have included values well below zero. Thus, the actual Zn concentration probably lies closer to 9 lg g À1 for LAP 031047 and this upper bound has been used in Fig. 5b. Fig. 5 shows CI and Sc normalized abundances of the 47 elements arranged in order of increasing putative volatility and grouped with cosmochemically similar character. We used Sc for normalization because of a lack of appropriate major elements (such as Si or Mg) to compensate for volatile content. Because of the volatility-controlled trends within LAP 031047, if we had used any other element than Sc as a normalizing element, our conclusions would remain unchanged. The element groups include lithophile (n = 25), siderophile (n = 7) and moderately volatile (n = 13) elements. Additionally, to accommodate ease of comparison, Table 4 shows the mean abundances for these element groups. Table 4). See Section 3.4 for normalization discussion. Other chondrite data from Friedrich et al. (2003Friedrich et al. ( , 2004 and Troiano et al. (2010).

Lithophile abundances within ordinary chondrites are known to be similar across all the ordinary chondrite groups (Kallemeyn et al., 1989), and LAP 031047 possesses lithophile abundances akin to other ordinary chondrites ( Table 4). Note that the H chondrite value, which is based on only two samples, is likely low. We suspect LAP 031047 will match the lithophile abundance of the L and LL chondrites as additional samples are added to our comparison database. Siderophile abundances lie between the H and L chemical groups but moderately volatile and thermally labile elements (such as Zn, Sn, Te, Se, and Cs) are severely depleted in LAP 031047 compared to average concentrations in ordinary chondrites (Fig. 5, Table 4). In fact, the degree of depletion seen in LAP 031047 is among the most extreme we have encountered within our studies of ordinary chondrite compositions. The smooth depletion trends seen in LAP 031047 indicate volatility-based (cosmochemical) rather than crystal-chemical (geochemical) control of the observed depletion.

Petrography

Macroscopic observations

The specimen is very friable and readily disaggregated to sand-size particles during sampling. It has a high primary porosity and a Medium Dark Gray (N4) color. However, a domain that has a sharp contact to the N4 domain is Medium Gray (N5) in color. Three coherent fragments between 1 and 2 cm in size are retained (Fig. 1). Tiny metal particles are visible on fresh broken surfaces and a thin dark fusion crust is visible on all these fragments.

Thin sections LAP 031047,4 and ,9 captured parts of a $100 lm thick fusion crust (Fig. 2) and many voids in the interior of the rock, which was also noted in the initial description (Satterwhite and Righter, 2006). This shows that high porosity is a property of the bulk rock and not just a localized phenomenon in the thin sections. Essentially, the rock appears to be composed of crushed chondrules.

Synchrotron X-ray microtomographic images of LAP 031047 show metal particles concentrated as rounded to vermicular nuggets in rounded void spaces in-between silicate domains (Fig. 6). Occasionally, traces of (light-gray) sulfide particles extend from the metal particles into surrounding silicates. Round to sub-rounded chondrules, some of which appear fragmented with round outlines that fade into irregular shapes, are visible in the tomographic image sequence.

Chondrules

Thirteen well recognizable, sometimes fragmented chondrules in thin section LAP 031047,4 have maximum apparent sizes between 0.4 and 1.48 mm (average: 0.86 mm) (Fig. 7). According to Brearley and Jones (1998), these chondrule sizes are typical for L and LL chondrites (0.7 and 0.9 mm average sizes, respectively), while H chondrite chondrules are only 0.3 mm in diameter, on average; however, our average is based on a small number of chondrules and is notably larger than the average apparent diameter of 0.57 mm Nelson and Rubin (2002) determined for type 3 LL chondrites. Chondrule types are porphyritic olivine-pyroxene (3), porphyritic pyroxene (4), porphyritic olivine (3), radial pyroxene (1), granular olivine (1), and a possible metal chondrule (1). In these igneous chondrules, $20-100 lm size phenocrysts are embedded in brownish relic melt that is transformed to microcrystalline, birefringent plagioclase.

Mineral clasts

Single mineral clasts of olivine are common and can be larger than 1 mm. Low-Ca pyroxene is less frequent but also occurs in single mineral clasts up to 1 mm, while high-Ca pyroxene is rare. Euhedral chromian spinel occasionally occurs associated with small olivine fragments and as components of porphyritic olivine-pyroxene chondrules. For example, a 0.5 mm olivine clast (average Fa composition of 21.4 mol%) has random intergrowths of typically 5-10 lm-size euhedral chromian spinel crystals (Fig. 8). Similar to other silicate minerals in LAP 031047, this olivine-rich clast contains linear arrays of troilite and minor Fe-Ni metal inclusions that manifest shock blackening effects. One anomalously large, 5-10 lm inclusion in the olivine is an intergrowth of troilite and Fe-Ni metal with plagioclase and some minor pyroxene and likely represents a crystallized melt inclusion. The surrounding matrix components contain some partly oxidized troilite and metal aggregates.

Metal-sulfide

Larger metal grains in LAP 031047 have sometimes spherical shapes and are surrounded by voids. Other shapes are elongated, branching masses between larger silicate grains. Associations with minor euhedral to anhedral sulfide particles occur occasionally. Sulfides commonly form tiny inclusions in silicates, as dispersed but also elongated blebs along crystallographic planes. Some sulfide is concentrated in veins between silicates. In these veins, micro-vesicular textures are evident. Metal particles typically exhibit domains of $5 lm size, euhedral, high Ni-Fe Metal (possibly taenite-martensite) in a low-Ni-Fe-metal matrix (possibly kamacite) (Fig. 9). Several of these metal aggregates are oxidized but such alteration is not widespread.

Shock petrography

The many fractures in olivine clasts have irregular to sub-parallel and sometimes radial orientations. Radial fractures originate from point sources on the rims of mineral clasts (Fig. 10). Such features were described by Kieffer (1971) from shocked sandstone in Meteor Crater and termed "concussion fractures". Another ubiquitous feature in LAP 031047 that relates to shock metamorphism is the blackening of silicate grains due to sub-lm metal and troilite particles (Fig. 10). These particles are frequently aligned in curved, linear directions, and planes, but also occur as chaotic dusting.

Following the shock classification scheme of Stö ffler et al. (1991), 20 randomly selected olivine grains larger than $50 lm were examined in thin section LAP 031047,4. No opaque shock veins, irregular melt veins, isolated melt pockets, or melt dikes occur, any of which would indicate shock stage S3 or higher (Stö ffler et al., 1991). Nine of the twenty olivine grains appear unshocked, with angles of rotation for complete extinction of 0-2°. The remaining eleven grains exhibit angles of rotation for complete extinction between 3°and 8°. Seven grains display a single set of multiple, parallel fractures and an additional two grains present two sets of multiple, parallel fractures, which is indicative of S2-3 conditions. Only one olivine grain was found to exhibit crystal-plastic deformation reminiscent of mosaicism; otherwise, no mosaicism, indications for planar deformation features or recrystallization occur. In summary, 45% of olivine grains appear unshocked (S1), 40% very weakly shocked (S2, up to 5-10 GPa) and 15% weakly shocked (S3) to a maximum of 15-20 GPa with an associated estimated temperature increase of up to 150°C (Stö ffler et al., 1991). We conclude that LAP 031047 records an average shock metamorphic overprint <10 GPa (S2).

Compositions of mineral components

3.6.1. Olivine

Chemical zonations are subtle and occur within larger grains that have FeO-enriched rims. Occasionally, olivine is intergrown with sub-lm size chromite, sulfide and Fe-Ni metal grains. Olivine compositions ; n = 138) are in between the typical ranges of H (Fa 16-20 ) and L chondrites (Fa 22-26 ), (Brearley and Jones, 1998) (Fig. 11a; Table 5).

Pyroxene

Occasionally, low-Ca pyroxene is compositionally zoned with respect to MgO and FeO. Stoichiometric low-Ca pyroxene ) is the dominant pyroxene in LAP 031047 (Table 6) and its compositions mostly fall in the range of H chondrites (Fig. 11b). Although low-Ca pyroxene mostly plots within the typical range of type 4-6 ordinary chondrites (Fig. 12), it appears to be richer in CaO than is typical for low-Ca pyroxene of type 4-6 ordinary chondrites (Brearley and Jones, 1998). This is expressed by the occurrence of pigeonite (Table 6). High Ca-pyroxene was found as overgrowth rims on low-Ca pyroxene crystals and as an inclusion in olivine.

The two-pyroxene thermometer of Lindsley and Anderson (1983) was used to constrain the equilibration temperature for LAP 031047. By using the QUILF program of Anderson et al. (1993), a best fit for enstatite-pigeonite analysis pairs yields a model temperature of $1100 to $1150°C at 1 bar pressure. No satisfactory thermometric model could be produced for the four augite analyses with pigeonite or enstatite data of LAP 031047.

Feldspar

Stoichiometric feldspar in LAP 031047 has a wide range of andesine to labradorite compositions (Ab 34-64 , An 33-64 , Or 0.4-2.8 ; Fig. 13; Table 7). This is unusual for type 4-6 ordinary chondrites (Brearley and Jones, 1998, and references therein), yet comparable to the range of compositions reported for H4 chondrite Yamato 74155 (Nagahara, 1980). Some material in the interstitial chondrule matrix appears turbid but the microprobe data did not confirm non-stoichiometric, alkali-and SiO 2 -rich materials (e.g., Gomes and Keil, 1980), which suggests that all chondrule melt mesostasis in LAP 031047 crystallized to plagioclase.

Fe-Ni metal

A bulk content of 7.9 wt% Ni is indicated by the average of 316 measurement points across 8 metal particles in LAP 031047. The dominant metallic mineral is kamacite with average Ni and Co concentrations of 6.50 and 0.59 wt%, respectively (Table 8). This average Ni concentration shows an affinity towards equilibrated L chondrites (Brearley and Jones, 1998) but the Co concentration values bridge the range of values that are characteristic for equilibrated H and L chondrites (Afiattalab and Wasson, 1980;Kallemeyn et al., 1989;Rubin, 1990). This indicates an intermediate composition between H and L chondrites of the kamacite component in metal particles of LAP 031047.

Taenite-martensite exhibits a range of Ni concentrations from 7.6 to 29.4 wt% ( Table 8) and low median P concentrations <0.1 wt% (Table 8), which suggests P under-saturation. However, four measurements indicate P concentrations between 0.61 and 1.76 wt%, possibly due to minute phosphide inclusions. Concentrations of Co in taenite-martensite with Ni-contents >25 wt% are between 0.18 and 0.26 wt% and follow the typical trend of equilibrated L and H chondrites (Afiattalab and Wasson, 1980). The relatively large scatter of the data does not allow clear distinction between L or H affinities.

Troilite

Two troilite analyses yielded a composition that contained 0.12 and 0.10 wt% Ni and 0.17 and 0.16 wt% Cr, while Mg, V, Co, Si, P, Ca, Ti, Cu, Mn, and Zn were <0.1 wt%.

Petrologic classification

Most petrologic features of LAP 031047 point towards a high degree of equilibration, possibly type 5 (van Schmus and Wood, 1967). This is manifested by: Orthorhombic low-Ca pyroxene; Typical feldspar grain size <50 lm; Chondrules, or fragments thereof, are readily delineated but not very well defined; No igneous glass was identified and the matrix is composed of voids and fragments of euhedral olivine and pyroxene with anhedral plagioclase in interstitial spaces. According to van Schmus and Wood (1967), unmetamorphosed chondrites contain only kamacite, while kamacite-taenite assemblages such as those of LAP 031047 require temperatures of $400-550°C.

Conflicting evidence to the assignment of petrologic type 5 for LAP 031047 is present with the compositions of low- Ca pyroxenes and olivines, and the shapes of metal particles. The mean standard deviations of FeO concentrations of 6.2% in the low-Ca pyroxenes, and 3.7% in olivines indicate poor equilibration of petrologic type 3-4 (van Schmus and Wood, 1967). This observation is at odds with the rounded to sub-rounded shapes of metal grains, which only agree with morphologies of such grains in type 6 ordinary chondrites (Afiattalab and Wasson, 1980). Alternatively, LAP 031047 was interpreted as an impact melt with rounded composite metal-sulfide particles (Satterwhite and Righter, 2006).

Classification

The initial classification as an L chondrite impact melt was based on the composition of olivine (Fa 23 ) and low-Ca pyroxene (Fs 17 ) and rounded metal aggregates. However, L chondrites are characterized by olivine and low-Ca pyroxene compositions of Fa 22-26 and Fs 19-22 , respectively (Brearley and Jones, 1998). Our electron microprobe data of LAP 031047 ( Fig. 11a and b; Tables 5 and 6) confirms the initially determined, intermediate compositional characteristics but shows a wider variability of Fa and Fs compositions. An H/L-chondritic character is also indicated by the oxygen isotopic composition ( Fig. 4 and Table 2) and Co concentrations of kamacites in LAP 031047, which scatter between the highest values for equilibrated H chondrites and the lowest values for equilibrated L chondrites (Table 8; Rubin, 1990). We also note that apart from the rounded shapes of metal aggregates, no petrographic evidence for impact melting is present in LAP 031047. Consequently, we conclude LAP 031047 is an H/ L type ordinary chondrite. Kallemeyn et al. (1989) (Clayton et al., 1991; Table 2). In this context, it is interesting to note that the oxygen isotopic compositions of Bremervö rde and Tieschitz are indistinguishable from those of L3 chondrites (Clayton et al., 1991). Unlike these two unequilibrated H/L chondrites, LAP 031047 exhibits siderophile element abundances that are typical for equilibrated ordinary chondrites ( Fig. 5; Table 4). The H/L4 Cali meteorite (Trigo-Rodríguez and Llorca, 2006) was described as exhibiting unequilibrated low-Ca pyroxene compositions with an average within the typical range for H chondrites, while its olivines have compositions that fall in the range for L chondrites (Brearley and Jones, 1998). Kamacite in Cali has Co concentrations of 0.66 wt%, similar to LAP 031047, and intermediate between equilibrated H and L chondrites. Trigo-Rodríguez and Llorca (2006) determined a porosity of 3.2 vol% for Cali that is much smaller than the $27 vol% porosity of LAP 031047. Also, thermoluminescence properties of Cali indicate a normal thermal and radiation history, suggesting the meteorite has not been within 0.95 AU of the Sun (Trigo- Rodríguez and Llorca, 2006). Unfortunately, comparison with oxygen isotopic signatures of equilibrated H/L chondrites is not possible because, to our knowledge, such data are not available. More petrographic and oxygen isotope data from the 39 H/L specimens that span almost the full spectrum of ordinary chondrite petrologic types (Meteoritic Bulletin Database, accessed April 7, 2011, http://tin.er.usgs.gov/meteor/metbull.php) are required to characterize this intermediate compositional group of meteorites.

Comparison with other H/L chondrites

Thermometric constraints

The difficult assignment of a petrologic type for LAP 031047 ("unequilibrated type 5") could be explained with a transient heating pulse, which briefly established petrographic type 5 conditions of 700-750°C (Dodd, 1981) on LAP 031047. The prevalence of orthopyroxene among low-Ca pyroxenes confirms heating above 630°C, the temperature at which low-Ca clinopyroxene transforms to orthopyroxene (Boyd and England, 1965). The much smaller diffusion coefficients in pyroxene and olivine than silicate glass and the lower activation energies required to recrystallize the glass (e.g., Brady, 1995) probably governed formation of the "unequilibrated type 5" assemblage. Reheating did not fully homogenize the original compositional zoning in pyroxenes and olivines that was established during the formation of chondrules at cooling rates between 10-1000°C/h (e.g., Hewins et al., 2005). Consequently, the reheating event must have been shorter than cooling from peak thermal metamorphic conditions for the type 4-6 ordinary chondrites, which would have otherwise equilibrated the pyroxenes and olivines in LAP 031047. For meter-size blocks of chondritic material, Rubin et al. (2008) suggested that at peak metamorphic conditions of $950°C, diffusional homogenization of Fe and Mg would take on the order of 1 year. This estimate also grossly agrees with the cooling rates of degrees per day at temperatures of 500-600°C in an H5 chondrite that Molin et al. (1994) inferred from disordering of Fe and Mg in M1 and M2 sites of pyroxenes. Thus, we note that the re-heating to 700-750°C that affected LAP 031047 likely lasted on the order of weeks rather than years. Moreover, reheating within the last $100 Ma was sufficient to degas the 40 Ar from LAP 031047 (Fig. 14).

Shock petrographic implications: LAP 031047 exhibits shock blackening, concussion fractures, high porosity of $27 vol%, and deformation features in olivine indicative of an average shock metamorphic overprint <10 GPa (S2) and associated average post-shock temperature increase of $50°C (e.g., Stö ffler et al., 1991). Abundant concussion fractures form in particulate olivine samples during pore collapse in a low-pressure regime (Bauer, 1979), which was further constrained by Kieffer (1971) to $5.5 GPa based on quartz of Meteor Crater sandstone target rocks. Under these low shock pressure conditions, mechanical effects dominate the response of powdered ordinary chondritic materials, while melting only becomes significant at shock pressures >25 GPa (Bauer, 1979;Hö rz et al., 2005). Hö rz et al. (2005) also demonstrated that shock pressures of 14.5 GPa close all pore space in L6 chondrite powders. Shock blackening of ordinary chondrites is typically related to shock pressures >5-10 GPa (S3) and higher (e.g., Stö ffler et al., 1991;Rubin, 1992;Britt and Pieters, 1994) although van der Bogert et al. (2003) and Consolmagno and Britt (2004) pointed out that shock blackening features are also reported in S1 and S2 ordinary chondrites. Consolmagno and Britt (2004) further suggest that the degree of shock metamorphic overprint correlates more strongly with the reduction of porosity than with shock blackening. It is noteworthy that the small troilite and metal inclusion trails in silicate grains of LAP 031047 likely resulted from the annealing of sulfide and metal injections (Rubin, 1992). Consequently, heating could have partially erased prior shock metamorphic features like mosaicism, undulous extinction, and crystallized aphanitic melt veins (Rubin, 1992). However, the high volume of retained porosity in LAP 031047 is inconsistent with impact-induced heating as a by-product of shock metamorphism. The rounded pores in LAP 031047 are also unlike the angular spaces that are commonly interpreted as porosity due to incomplete compaction (e.g., Friedrich et al., 2008;Sasso et al., 2009). Because LAP 031047 suffered only minor weathering, and its porosity does not primarily reside in fractures, increased porosity due to freeze-thaw cycling in Antarctica (Consolmagno et al., 1998) is unlikely. This suggests that a different process, possibly sintering of regolith on its parent asteroid, may have produced LAP 031047's petrofabric.

Metallography: The abundance of Co in kamacite of LAP 031047 confirms an intermediate character between L and H chondrites. Metal particles in LAP 031047 do not contain tetrataenite (e.g., Yang et al., 1996) and there are, thus, no "unzoned plessite" particles (Reisener and Goldstein, 2003). Comparable analogies to the metallographic textures in LAP 031047 are the coarse, polycrystalline kamacite and taenite intergrowths that were described by Heymann (1970, 1971) and Bennett and McSween (1996) in the Kingfisher L5 chondrite and the Sweetwater H5 chondrite. According to these authors, the Kingfisher and Sweetwater chondrites experienced reheating events, which transformed their taenite-martensite particles to, on average, 11 and 2.5 lm size taenite crystals in kamacite matrices, respectively. Taylor and Heymann (1971) assigned a post-shock temperature increase of 600°C for such textures in Kingfisher and 480-570°C for Sweetwater. This is based on the presumption that reheating of martensitic Fe-Ni metal into the binary a-c field induces rapid decomposition into kamacite and taenite. If the metallographic textures of LAP 031047 resulted from a similar massive transformation of martensite or taenite with an average Ni concentration of 7.9 wt% to kamacite and taenite, they record cooling from reheating to at least 700-750°C (e.g., Wood, 1967). The central taenite composition method constrains the relative metallographic cooling rate of LAP 031047 to much faster than 100°C/Ma, because of the small high-Ni particle sizes and corresponding Ni concentrations <30 wt% (Willis and Goldstein, 1981). The analogy with Sweetwater and Kingfisher is limited, though. It is noteworthy that both of these meteorites contain relics of zoned martensitic textures, which were only partly transformed to kamacite and taenite. The metal in LAP 031047 appears to have formed from unzoned taenite particles, suggesting some equilibra-tion of the precursor at high temperatures before development of the kamacite/taenite-martensite texture.

Trace element geochemistry: For the cause of depletion of moderately volatile and thermally labile elements (Ga to Zn in Fig. 5), we can rule out a removal of Fe-FeS from the eutectic assemblage because refractory lithophiles, which if anything are depleted in LAP 031047, are not enriched enough for significant volatile loss to have occurred. Additionally, siderophile abundances seem to bear a resemblance to H chondrites, so they have probably not been lost. One possible explanation for the depletion of moderately volatile and thermally labile elements is that LAP 031047 accreted at a higher temperature than H chondrites, so that volatiles were never incorporated in the first place. Another possibility is the loss of volatiles during an impact scenario, but this surely would have increased the lithophile content significantly given mass balance considerations. Also, it is conceivable that thermally labile elements were lost due to sublimation during heating in a near-surface, low-gravity environment, when metal particles re-equilibrated and silicates annealed. In agreement with the depletion trends in LAP 031047, heating experiments e.g., of Allende (Lipschutz and Woolum, 1988, and references therein) and the Tieschitz H/L3.6 chondrite (Ikramuddin et al., 1977) constrained the onset of the loss of the thermally labile trace elements from Zn to Ga to $600-1000°C. Moreover, Lipschutz and Woolum (1988) suggest the loss of noble gas nuclides in that temperature range.

Noble gas inferences

Other ordinary chondrites with similarly young Ar-Ar ages are H5 Ucera (40 ± 30 Ma; Bogard et al., 1976) and L6 Clovis (No. 2) (<40 Ma; Turner and Cadogan, 1973). Interestingly, L5 Farmington has an Ar-Ar degassing age of 510 ± 40 Ma (Bogard and Hirsch, 1980), but one of the youngest cosmic-ray exposure ages of $0.035 Ma (Patzer et al., 1999). H4 Seres also has a very young CRE age of 0.64 Ma (Heymann et al., 1967;Graf and Marti, 1995), and like Farmington, has been reheated (Wood, 1967). This author also suggested that Seres may have been in a pre-terrestrial orbit very close to the Sun, which caused a late heating event to several hundred°C, in order to explain the metallographic texture and depletion of cosmogenic noble gases. A similar conclusion has been postulated for Farmington to explain its very young CRE age (e.g., Heymann et al., 1967;Oberst, 1989). Our constraints for a very young CRE age <0.5 Ma imply LAP 031047 was shielded from galactic cosmic rays before $0.5 Ma, possibly because its location was >1-2 m from the surface of its parent asteroid (e.g., Eugster, 2003;Leya and Masarik, 2009) or it was reheated in the last 0.5 Ma. The pervasive degassing of LAP 031047 at $100 Ma ago implies two possible sources for heating: (1) waste-heat from shock metamorphism and (2) solar heating. As outlined in the previous sections, the pervasive degassing and the high porosity of LAP 031047 discredit scenario (1). Although passive heating due to close proximity with hot impactites may still offer a viable explanation of the major petrologic characteristics of LAP 031047, the plausibility of scenario (2) requires scrutiny.

Near-Earth objects

Analogous to Farmington and Seres, LAP 031047 could have belonged to a Near-Earth Object (NEO) with a small perihelion distance to the Sun. Dynamic lifetimes of NEOs are estimated to be on the order of 10-100 Ma (Gladman et al., 2000, and references therein), which agrees with chronological constraints for LAP 031047. The major group of objects with Earth-crossing orbits is the Apollos ($62% of the predicted population of NEOs; Bottke et al., 2002), which have semi-major axes greater than that of the Earth (>1 AU) and perihelion distances <1.017 AU. The less populous Atens have semi-major axes <1 AU, so they are usually inside the orbit of Earth. Both NEO groups can experience close encounters with the Sun and, consequently, suffer substantial heating, although Atens are more likely to do so. A prominent Earth-crossing object belonging to the Apollos is 3200 Phaethon.

Phaethon

3200 Phaethon is a 5-8 km diameter Apollo with a perihelion distance of 0.14 AU, which makes it a near-Sun object as well (e.g., Ohtsuka et al., 2009). According to Bottke et al.'s (2002) steady-state orbital distribution model of NEOs, Phaethon has a 20% chance of coming from the central main belt and an 80% chance of coming from the inner main belt, making it a plausible ordinary chondrite parent. Phaethon is the source for the Geminid meteoroids, which have average velocities of $36 km/s. Although these average velocities are higher than those for most near-Earth asteroids (average 18 km/s; Chyba, 1991), they are fully consistent with a main belt origin for Phaethon, because the latter developed an unusually large eccentricity and inclination. The Geminid meteoroids also have a dynamic strength that is substantially higher than the material delivered from most other meteoroid streams, which are usually linked with cometary parent bodies (Trigo-Rodríguez and Llorca, 2006). According to Babadzhanov (2002), Geminid meteoroids have a density of 2.9 ± 0.6 g/cm 3 , which falls in the typical range of ordinary chondrites (Consolmagno et al., 2008;Opeil et al., 2010) and the density of LAP 031047 (2.69 ± 0.06 g/cm 3 ). The only other meteoroid material with a comparably high density in Babadzhanov's (2002) study stems from the d Aquarids (2.4 ± 0.6 g/cm 3 ), which have a substantially lower average strength than the Geminid meteoroids (Trigo-Rodríguez and Llorca, 2006). Phaethon is spectrally classified as an F or B-type asteroid, which are interpreted as dehydrated CI and CM carbonaceous chondrites that underwent thermal processing (e.g., Hiroi et al., 1996;Licandro et al., 2007). Thermal modeling of Ohtsuka et al. (2009) for Phaethon infers subsolar (the location on the surface that sees the Sun at its zenith) temperatures of 527-827°C at its perihelion, which induces temperatures <400°C over parts of its surface for 1-3 weeks during its perihelion passage. In order to explain diverse Na abundances in Geminid meteors, Kasuga (2009) inferred thermal processing at temperatures of 900 K ($627°C) on the surface of Phaethon, which caused sublimation of Na from alkali silicates. In analogy with Mercury, which has a subsolar temperature at perihelion of 450°C (Vilas, 1988), thermal vaporization, photon stimulated desorption, and ion-sputtering (e.g., Killen et al., 2004) should also cause Na and K loss on Phaethon's surface.

Solar heating

Basic constraints: In order to estimate the heating processes near the surface of a Phaethon-like object, we use parameters from Ohtsuka et al. (2009) and a fast rotating black-body approximation for the behavior of the asteroid. In agreement with Ohtsuka et al.'s (2009) estimate, this gives surface temperatures at perihelion between 470 and 780°C. However, a first order estimation of the thermal diurnal skin depth (Harris and Lagerros, 2002) of this object is on the order of 0.1 m. We would therefore expect only the uppermost surface of this object to be severely heated.

Penetration depth of solar heating: A more elaborate approach to simulating the solar heating of subsurfaces of planets and small bodies is based on a uni-dimensional finite difference model described in Lasue et al. (2008) that has been adapted for a rocky asteroid. This model takes the diurnal temperature variation at the surface of the asteroid into account. Fig. 15a illustrates modeled temperature profiles as a function of the depth in the near subsurface of a Phaethon-like spherical body at perihelion below the subsolar point. Four profiles are shown along with the 700°C isotherm (solid horizontal line): The plain line designates the temperature profile under the subsolar point at the equator for zero obliquity, and the long dashed line indicates the temperature profile at the pole with a maximal obliquity of 90°for the same rocky asteroid. These profiles show that the penetration of the heat wave is very limited, in a way that practically no material in the subsurface gets heated above 700°C by the Sun. This is true even in the case of 90°obliquity, for which the asteroid's pole is directed towards the Sun at perihelion and no diurnal effect exists. Under these conditions of illumination for a prolonged time, maximum heating to a depth <0.1 m occurs in these regions of the asteroid. The effect of the heat wave in a regolithic material of low density (0.5 g cm À3 ) and low thermal conductivity (0.05 W m À1 K À1 ; Helbert and Benkhoff, 2003) is displayed by the dotted and dashed dotted lines. Because this material conducts heat very poorly, the penetration of the heat wave is depressed to even lower depths. This suggests that a cm-thick layer of regolith would effectively insulate the subsurface from heating above $700°C by diurnal effects during passage of its parent asteroid through the perihelion even in the case of high-obliquity illumination.

Extremes beyond Phaethon: If our assumptions for the thermal properties of 3200 Phaethon's subsurface are correct, it appears unlikely that solar heating on a Phaethonlike near-Sun object could produce the heating effects and petrologic characteristics recorded by LAP 031047. Nonetheless, other small bodies exist that have even smaller perihelion distances to the Sun. The Jet Propulsion Laboratory's Small Bodies Database search engine lists 41 currently known objects that have perihelion distances <0.2 AU. The closest, 2005 HC 4 , squeezes past the Sun at 0.071 AU, but the most interesting may be 2008 FF 5 , which has a perihelion distance of 0.079 AU, a semi-major axis of 2.277 AU, and inclination of orbit of 2.62°. Essentially, this means 2008 FF 5 crosses the Earth's orbit twice per rotation around the Sun, giving it an impact probability with Earth of 3.6 Â 10 À7 (NASA Near-Earth Object Program, http:// neo.jpl.nasa.gov/risks/2008ff5.html). Assuming the physical characteristics of 3200 Phaethon, 2008 FF 5 's surface is heated above 1100°C. Fig. 15b shows the temperature profile in the subsurface of 2008 FF 5 for a subsolar temper-ature of 1100°C and the physical properties used for Phaethon in Fig. 15a. It can be seen that only material located a few cm below the surface is heated above 700°C in the case that does not take obliquity into account. In the most favorable case of the north pole facing the Sun at perihelion with an obliquity of 90°, heating above 700°C penetrates to maximum depths of $30 cm. Again, the presence of a regolith layer with a minimal thickness of 2 cm would effectively insulate the subsurface from heating above 700°C, if the obliquity is equal to 0°. In the most favorable case of high obliquity, the regolith and surface material is going to suffer heating to 700°C down to 20 cm.

Solar heating of LAP 031047 ?

Pending further analyses of products from the interaction with solar wind and cosmic rays, we are faced with a conundrum because neither impact-related heating nor solar heating can explain the petrologic characteristics of LAP 031047 comprehensively.

(1) The solar heating scenario seems improbable because it would require that LAP 031047 was exposed at the surface, e.g., as a boulder, on a near-Sun object with a high obliquity that allowed prolonged solar heating, which degassed LAP 031047. Then, $100 Ma ago, LAP 031047 was shielded from the solar-wind and galactic cosmic-rays, implying burial to a depth >1-2 m (Eugster, 2003;Leya and Masarik, 2009). (2) Alternatively, pervasive degassing and thermal annealing at temperatures >$700°C, $100 Ma ago could have been caused by (passive) heating in close proximity to impact melt at a depth >1-2 m.

Both scenarios then require an impact event <0.5 Ma ago, which ejected LAP 031047 as part of a meter-size object that was exposed to cosmic rays.

CONCLUSIONS

Antarctic meteorite LAP 031047 is an ordinary chondrite with intermediate chemical characteristics between H and L chondrites. The intermediate nature is expressed by its oxygen isotope, major mineral, and lithophile and siderophile element composition. A striking characteristic is its high porosity of $27 vol% and apparent low shock metamorphic overprint of <10 GPa. Although olivine and low-Ca pyroxene are not equilibrated, petrologic type 5 metamorphic conditions best describe LAP 031047's petrographic characteristics. Another peculiar feature is the metallographic texture, which likely formed from massive transformation of taenite to kamacite due to reheating above $700°C. Pervasive reheating to high temperatures is also indicated by strong depletion of moderately volatile trace elements and LAP 031047's Ar-Ar radioisotopic age of 100 ± 55 Ma. As cause for reheating, only (1) passive heating from hot impactites, or (2) solar heating, appear plausible. The theoretical possibility of delivery of material from near-Sun objects is manifested by the examples of 3200 Phaethon and several small bodies that experience solar heating in close perihelion distances to the Sun. Fig. 15. Finite-difference modeling of near-Sun objects. Phaethonlike objects have been modeled with and without obliquity as spherical bodies with a rotation period of 3.6 h, an albedo of 0.1, and a diameter of 5.1 km (see Ohtsuka et al., 2009, and references therein). A bulk density of 2.69 g cm À3 with a thermal conductivity equal to 1.5 W m À1 K À1 is used. This thermal conductivity value is adapted from Opeil et al. (2010), who studied the physical properties of the L chondrite Lumpkin and the CK4 chondrite NWA 5515, which have similar densities and porosities than LAP 031047. A heat capacity of 800 J kg À1 K À1 is used that is typical for meteorites (see e.g., Opeil et al., 2010). An about one order of magnitude lower thermal conductivity and density than the meteoritic material is used for the simulation of regolithic material (Helbert and Benkhoff, 2003). (a) Temperature as a function of the subsurface depth for an asteroidal body in the current orbit of 3200 Phaethon: the solid horizontal line designates the 700°C isotherm; other lines display temperature profiles for meteoritic and regolithic material below the subsolar point at perihelion at the equator of the body for zero obliquity and near the illuminated pole for 90°o bliquity. (b) Temperature as a function of the subsurface depth for an asteroidal body with a perihelion distance of 0.079 AU, alike 2008FF 5 ; same signatures as in Fig. 15a.

However, such heating is limited to the uppermost $30 cm of the surface region of a near-Sun object even in the most favorable circumstances of maximum exposure to the Sun. This suggests that its young degassing age, and unusual petrographic and cosmochemical characteristics are more likely impact-related features due to secondary heating of LAP 031047 close to hot impactites $100 Ma ago. Its CRE-age indicates LAP 031047 became part of a meter-size asteroid <0.5 Ma ago, which was captured by Earth.

We anticipate that detailed studies of LAP 031047's cosmic-ray and solar wind exposure history could further unravel the evolution of its parent object.

Table 3

Table 5

Table 6