Collisional facilitation of aqueous alteration of CM and CV carbonaceous chondrites

Available online at www.sciencedirect.com

Geochimica et Cosmochimica Acta 90 (2012) 181–194 www.elsevier.com/locate/gca

Collisional facilitation of aqueous alteration of CM and CV carbonaceous chondrites Alan E. Rubin Institute of Geophysics and Planetary Physics, University of California, Los Angeles, CA 90095-1567, USA Received 8 December 2011; accepted in revised form 9 May 2012; available online 18 May 2012

Abstract CM chondrites exhibit a strong correlation between the degree of alteration and the extent of particle alignment (i.e., the strength of the petrofabric). It seems likely that the S1 shock stage of essentially every CM and the high matrix abundance (70 vol.%) of these samples ensured that the shock waves that produced CM petrofabrics (by collapsing matrix pores and squeezing chondrules into pore spaces) were significantly attenuated and were too weak to damage olivine crystal lattices. Random collisions on the CM body produced petrofabrics and created fractures in the target rocks. Subsequent impact-mobilization of water caused hydrated phases to form preferentially in the more-fractured regions (those with the strongest petrofabrics); the less-deformed, less-fractured CM regions experienced lower degrees of aqueous alteration. Many CV3 chondrites also have petrofabrics: roughly half are from the oxidized Bali-like subgroup (CV3OxB), roughly half are from the reduced subgroup (CV3R) and none is from the oxidized Allende-like subgroup (CV3OxA) (which is less altered than CV3OxB). Nearly all CVs with petrofabrics are S3–S4 and nearly all CVs that lack petrofabrics are S1. Oxidized CVs have much higher porosities (typically 20–28%) than reduced CVs (0.6–8%), facilitating more-extensive aqueous alteration. The CV3R chondrites formed from low-porosity material that inhibited oxidation during alteration. The oxidized CV subgroups formed from higher-porosity materials. The CV3OxB samples were shocked, became extensively fractured and developed petrofabrics; the CV3OxA samples were not shocked and never developed petrofabrics. When water was mobilized, both sets of porous CV chondrites became oxidized; the more-fractured CV3OxB subgroup was more severely altered. Ó 2012 Elsevier Ltd. All rights reserved.

1. INTRODUCTION CM carbonaceous chondrites experienced variable degrees of aqueous alteration on their parent body (McSween, 1979; Kerridge and Bunch, 1979; Bunch and Chang, 1980; Zolensky et al., 1997; Browning et al., 2000; Trigo-Rodrı´guez et al., 2006; Rubin et al., 2007). Some individual CM chondrites (e.g., Murray, Cold Bokkeveld) are breccias that contain sub-centimeter-size clasts that have been altered to different extents than their hosts (Rubin and Wasson, 1986; Metzler et al., 1992). The more-altered CM chondrite materials contain very low modal abundances of metallic

E-mail address: [email protected] 0016-7037/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.gca.2012.05.016

Fe–Ni, altered kamacite, high proportions of chondrules with altered mafic silicate phenocrysts, altered refractory inclusions, high phyllosilicate/(olivine + pyroxene) modal ratios, high Mg–serpentine/Fe–cronstedtite ratios, low amounts of pyrrhotite, diverse carbonate compositions, low modal abundances of “PCP” clumps (formerly dubbed “poorly characterized phases”), and “PCP” with low contents of oxidized iron (e.g., Greenwood et al., 1994; Rubin, 2007; Rubin et al., 2007; Howard et al., 2009, 2011; Beck et al., 2010; Palmer and Lauretta, 2011). (PCP consists of variable proportions of tochilinite, a layered Fe–S–O–Ni mineral, 2[Fe,Mg,Cu,Ni[ ]S]
1.57–1.85[(Mg,Fe,Ni,Al, Ca)(OH)2] (where the open bracket represents a vacancy) and the hydrated phase cronstedtite [Fe22+Fe3+(Si, Fe3+)2O5(OH)4]; Tomeoka and Buseck, 1985.)

182

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

The more-altered CM chondrites also appear to contain relatively high bulk H (reflecting abundant H2O), high D17O values (reflecting the heavy oxygen-isotopic composition of the water that altered these rocks) and low relative amounts of trapped planetary 36Ar (reflecting the destruction of retentive sites in the whole rock) (Browning et al., 1996; Clayton and Mayeda, 1999; Rubin et al., 2007). Temperatures during CM aqueous alteration have been estimated to range from 0 to 25 °C (e.g., Clayton and Mayeda, 1984; Rosenberg et al., 2001; Benedix et al., 2003), from 20 to 35 °C (Guo and Eiler, 2007), and up to 80 °C (Baker et al., 2002). CM falls contain 9 wt.% H2O+ (indigenous water; Jarosewich, 1990) bound as hydroxyl in phyllosilicates (e.g., Barber, 1981, 1985; Tomeoka and Buseck, 1985). Despite petrographic differences, most CM chondrites have similar modal abundances of hydrous phases (Howard et al., 2009). The water responsible for aqueous alteration in CM chondrites may have been derived from melted or vaporized ice (Grimm and McSween, 1989; Bland et al. 2009) or, possibly, from dehydrated phyllosilicates that had formed in the nebula (Wasson, 2008), perhaps during chondrule formation (e.g., Petaev and Wood, 1998; Wasson and Trigo-Rodrı´guez, 2004; Ciesla and Lauretta, 2005). Different schemes have been devised to assess the degree of aqueous alteration of CM chondrites (McSween, 1979; Browning et al., 1996; Hanowski and Brearley, 2001; Rubin et al., 2007; Howard et al., 2009; Beck et al., 2010). Recently, the scheme of Rubin et al. (2007) has been the most widely adopted because these authors used readily measurable mineralogical, petrological and textural parameters. In this method, the most-altered CM chondrites (previously classified as CM1) are designated CM2.0 (e.g., MET 01070; LAP 02277). These rocks have 60.02 vol.% metallic Fe–Ni and no remaining mafic silicate grains. Less-altered samples (e.g., CM2.5 Murchison, CM2.4/2.5 Murray) contain more metallic Fe–Ni and more-pristine mafic silicate grains. The least-altered sample studied by Rubin et al.

(2007) is CM2.6 QUE 97990 (Fig. 1a); it contains 0.2 vol.% metallic Fe–Ni (Table 1 of Howard et al. (2011)) and has unaltered mafic silicate grains. QUE 97990 also contains less magnetite than other CM chondrites (0.6 vol.% vs. 0.7–2.3 vol.%; Howard et al., 2011). The Paris CM chondrite appears to be even less altered than QUE 97990. It has experienced only mild thermal metamorphism and has low abundances of phyllosilicates and a high metal/magnetite modal abundance ratio (Bourot-Denise et al., 2010; Zanda et al., 2010). Hypothetical CM3.0 chondrites should have relatively abundant metallic Fe–Ni, pristine chondrules with glassy mesostases, no PCP clumps, and little or no phyllosilicates or carbonate. An important question is what caused some CM chondrites to become more aqueously altered than others. Young et al. (1999) suggested that down-temperature flow of an aqueous fluid within parent asteroids could be responsible for heterogeneities in the mineralogy and O-isotopic compositions of carbonaceous chondrites. The proximate cause of water mobilization on the CM body could have been gas pressure arising from the asteroid interior (e.g., Young et al., 1999), sporadic flow (e.g., Cohen and Coker, 2000; Coker and Cohen, 2001) or convection cells (e.g., McSween et al., 2002; Travis and Schubert, 2005). In general, the hydrological activity in these models is assumed to have been caused ultimately by the melting of accreted water ice heated by the decay of 26Al (e.g., Grimm and McSween, 1989; Young, 2001). Many CM chondrites are brecciated (e.g., Bischoff and Sto¨ffler, 1992; Metzler et al., 1992; Metzler, 2004; Bischoff et al., 2006) and contain solar-wind-implanted rare gases (Schultz and Kruse, 1989) indicating that they are regolith breccias that resided near the surface of their parent asteroid. Several CM2.0 chondrites have a foliation (e.g., Figs. 1b, 2, and 3) wherein chondrules and other components (e.g., chondrule mantles, phyllosilicates, PCP particles, sulfide grains) are approximately aligned (Fujimura et al., 1982; Zolensky et al., 1997; Rubin et al., 2007). Similar

Fig. 1. Back-scattered electron (BSE) mosaics of two CM chondrites showing the stark difference in the degree of preferential particle orientation. (a) QUE 97990 (CM2.6) exhibits essentially random particle orientations. (b) MET 01070 (CM2.0) shows a strong preferential orientation. A centimeter-size “PCP”-rich (cronstedtite–tochilinite) lens (white) transects the thin section. The lens is oriented in approximately the same direction as fractures in the whole rock and the long axes of many of the chondrule pseudomorphs. Both images are to the same scale.

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

Fig. 2. Portion of Cold Bokkeveld (CM2.2) showing alignment of altered chondrules (dark gray, center left and bottom left) and a large clump of PCP particles (light-gray clump at center to right) in a NE–SW direction. Not only is the PCP clump elongated in this direction, but so are the individual PCP particles. BSE image.

183

group (CV3R) that includes such meteorites as Leoville, Efremovka, Arch and Vigarano and is characterized by relatively high metallic Fe–Ni abundances, low-Ni metal, lowNi sulfide, minor magnetite and no phyllosilicates; (2) an Allende-like oxidized subgroup (CV3OxA) characterized mainly by chondrule primary minerals having been replaced by magnetite, Ni-rich sulfide, ferroan olivine (Fa 30–60), feldspathoids, andradite garnet, and kirschsteinite; and (3) a Bali-like oxidized subgroup (CV3OxB) characterized by abundant phyllosilicates and chondrules with primary minerals having been replaced by phyllosilicate, magnetite, Ni-rich sulfide, very ferroan olivine (Fa 95– 100), and hedenbergite (McSween, 1977; Weisberg et al., 1997; Krot et al., 1998, 2005). Of the two oxidized subgroups, the CV3OxB subgroup is the one that has been more significantly altered; it contains more matrix material, less metallic Fe–Ni, more magnetite, more-ferroan olivine, and more phyllosilicates than CV3OxA samples (e.g., McSween, 1977; Weisberg et al., 1997; Krot et al. 2005; Howard et al., 2010). All three CV subgroups have roughly similar distributions in their cosmic-ray exposure ages (Scherer and Schultz, 2000), consistent with them all being derived from the same parent asteroid. Because many CV chondrites have pronounced petrofabrics (Martin et al., 1975; Mu¨ller and Wlotzka, 1982; Kracher et al., 1985; Cain et al., 1986; Scott et al., 1992), it is worth exploring whether they exhibit a correlation between the degree of oxidation and the extent of structural deformation. 2. ANALYTICAL PROCEDURES

Fig. 3. Chondrule and particle alignment in CM2.0 LAP 02277. As an illustration, the orientation of the large chondrule pseudomorph in the image was measured by drawing a line parallel to the long axis of the particle, drawing a vertical line at its terminus and measuring the angle (using the measuring tool in Adobe Photoshop) between the two lines clockwise. BSE image.

foliations in ordinary and carbonaceous chondrites have been previously ascribed to shearing resulting from impacts (e.g., Sneyd et al., 1988; Gattacceca et al., 2005; Rubin and Swindle, 2011). The significant number of CM breccias and the prominent (plausibly impact-induced) petrofabrics evident in the highly altered CM2.0 samples raise the possibility that collisions could be in part responsible for the aqueous alteration of CM chondrites. That possibility is explored here. Another carbonaceous-chondrite group that has members with prominent petrofabrics is the CV group (e.g., Martin et al., 1975; Mu¨ller and Wlotzka, 1982; Kracher et al., 1985; Cain et al., 1986; Scott et al., 1992). This group has been divided into three subgroups: (1) a reduced sub-

Thin sections of nine CM chondrites (that had previously been classified into petrologic subtypes; Rubin et al., 2007) were studied microscopically in transmitted and reflected light with an Olympus BX60 petrographic microscope. The thin section numbers in this study are the same as those listed in Table 1 of Rubin et al. (2007). Also studied were three CV3 thin sections: MCY 05219,11, Gi-4/2 of NWA 7107 and Y 981208,51-2. Shock stages were assigned based on the criteria developed by Sto¨ffler et al. (1991). The standard meteorite abbreviations used here correspond to the following locations: ALH = Allan Hills; EET = Elephant Moraine; LAP = LaPaz Icefield; LON = Lonewolf Nunataks; MET = Meteorite Hills; MCY = MacKay Glacier; NWA = Northwest Africa; QUE = Queen Alexandra Range; SCO = Scott Glacier; Y = Yamato. I examined previously prepared mosaics of moderately high-resolution (2 lm/pixel) back-scattered electron (BSE) images of the CM thin sections and newly prepared transmitted-light images of the MCY 05219, NWA 7107 and Y 981208 sections. One section of each of the meteorites was studied. The BSE images were made with the LEO 1430 VP scanning electron microscope (SEM) at UCLA using a 15 keV accelerating voltage and a working distance of 26 mm. The transmitted light images of MCY 05219, NWA 7107 and Y 981208 were made with the Olympus petrographic microscope.

184

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

Each chondrule, chondrule fragment, refractory inclusion, amoeboid olivine inclusion, chondrule pseudomorph, and PCP particle or clump larger than 250 lm was individually numbered in the CM-chondrite mosaics and the images of MCY 05219, NWA 7107 and Y 981208. The “measure tool” device of Adobe Photoshop was used to compute the azimuth of the long axis (or longest direction in plan view) of each particle, measured clockwise from the top of each image (arbitrarily designated north) (e.g., Fig. 3). The median azimuthal angle was calculated for each meteorite and the deviation of the azimuth of the long axis of each object from the median azimuth was determined. The compositions of silicates and oxides in MCY 05219 and NWA 7107 were measured on the UCLA JEOL electron microprobe using natural and synthetic standards, an accelerating voltage of 15 keV, a sample current of 15 nA, 20-s counting times per element, a focused beam and ZAF corrections. 3. RESULTS 3.1. Shock stages of CM chondrites The vast majority of olivine grains in most of the CM chondrites in this study (QUE 93005, CM2.1; Cold Bokkeveld, CM2.2; QUE 99355, CM2.3; Y 791198, CM2.4; Murray, CM2.4/2.5; QUE 97990, CM2.6) exhibit sharp optical extinction under the microscope and are designated shockstage S1 (i.e., essentially unshocked; Sto¨ffler et al., 1991). Murchison (CM2.5) has a few olivine grains that show undulose extinction and many olivine grains with sharp extinction; it was classified as an S1–S2 breccia by Scott et al. (1992). However, because my observations indicate that fewer than 25% of the olivine grains are characteristic of shock-stage S2, I designate the Murchison whole rock as S1. Tomeoka et al. (1999) found that in their sample of Murchison, 20 of 22 olivine grains exhibited sharp optical extinction, and thus also classified Murchison as S1. There are no mafic silicate grains remaining in the highly altered CM2.0 chondrites LAP 02277 and MET 01070, so their shock stages cannot be determined. 3.2. Description of MCY 05219 MCY 05219 (CV3) contains abundant magnetite and Ni-bearing sulfide and very little metallic Fe–Ni, indicating that it is probably a member of the Bali-like oxidized subgroup (CV3OxB). Consistent with this classification is the presence in the matrix of abundant fayalitic olivine and accessory hedenbergite (Fs45Wo50; n = 2); no nepheline or sodalite was identified. Other members of the Bali subgroup include Bali and Kaba as well as portions of Mokoia, Grosnaja and ALH 85006 (Weisberg et al., 1997; Krot et al., 1998). Chondrules in MCY 05219 include porphyritic olivine (PO), porphyritic olivine-pyroxene (POP), porphyritic pyroxene-olivine (PPO), and barred olivine (BO) textural types; some are surrounded by igneous rims (Rubin, 1984; Krot and Wasson, 1995). A few of the pyroxene-bearing chondrules contain polysynthetically twinned low-Ca clinopyroxene. The chondrules average 800–900 lm in

Fig. 4. NWA 7107 thin section showing that chondrules and inclusions exhibit a preferred NW–SE orientation. Image courtesy of A. Greshake. Transmitted light.

apparent diameter, characteristic of CV3 chondrites (Rubin, 2010). Amoeboid olivine inclusions (AOIs) constitute about 5 vol.% of the available thin section. The rock exhibits a strong petrofabric: the chondrules, chondrule fragments, AOIs and sulfide–magnetite assemblages are all aligned in the same general direction. The porphyritic chondrules have been deformed and many have aspect ratios ranging from 1.5 to 2.5. The aspect ratios of the AOIs range from 1.6 to 4.0. Nearly half of the coarse olivine grains exhibit undulose extinction. Many of the mafic silicates have extensive irregular and planar fractures, indicative of shock-stage S3. A few rare grains exhibit weak mosaic extinction. 3.3. Description of NWA 7107 NWA 7107 (CV3) contains very heterogeneous olivine, ranging in composition from Fa 0.7 to 40.6 mol% (n = 43) with a mean value of Fa 31.7 ± 12.7. Low-Ca clinopyroxene (Fs 2.4 Wo 0.7) with polysynthetic twins and diopside (Fs 0.6 Wo 52) with significant Al2O3 also occur. Opaque phases include troilite, Ni-rich metal, and Ti- and Cr-bearing oxide grains; the apparent absence of magnetite indicates that the rock is a member of the reduced CV subgroup (CV3R). Chondrules in NWA 7107 average 900–1000 lm in apparent diameter, within the CV3 chondrule size range (Rubin, 2010). Chondrule textural types include PO, POP, PPO, PP (porphyritic pyroxene) and BO varieties; many of the chondrules have thick igneous rims. Amoeboid olivine inclusions constitute 8 vol.% of the thin section. The meteorite has a strong petrofabric, with chondrules, chondrule fragments and inclusions aligned in the same general direction (Fig. 4). Many of the porphyritic chondrules have been deformed; their aspect ratios range from 1.2 to 1.7. The aspect ratios of the AOIs range from 2.4 to 5.4. Most coarse olivine grains in NWA 7107 have planar fractures and exhibit weak mosaic extinction, indicative of shock-stage S4. 3.4. Description of Y 981208 Y 981208 (CV3) contains heterogeneous olivine (Fa 0.2– 32.0) and low-Ca pyroxene (0.4–5.8) with polysynthetic twins (Meteorite Bulletin Database; this study). Chondrule textural types include PO, POP, PPO, PP, BO and C

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

(cryptocrystalline) varieties; the mean apparent chondrule diameter is 870 ± 520 lm (n = 22). Some of the chondrules have igneous rims. The rock contains relatively abundant metal and sulfide and little magnetite, indicating that it is a member of the reduced subgroup (CV3R). Olivines exhibit weak mosaic extinction characteristic of shock-stage S4. The rock has a pronounced petrofabric; the aspect ratios of AOIs range up to 3.5. 3.5. Petrofabrics The degree of orientation of preferentially oriented particles in a randomly sliced thin section is a lower limit of the true degree of alignment because the sample is almost always oriented in a direction that is not exactly parallel to the maximum particle elongation. Thus, the degree of orientation of the chondrite particles reported here would be pushed toward higher values if the thin sections had been made intentionally to exhibit maximum alignment. MET 01070 (CM2.0) contains elongated PCP-rich lenses (0.06–1  8–12 mm) that are roughly aligned in the same direction as the long axes of the elliptical phyllosilicate-rich chondrule pseudomorphs (e.g., Fig. 1b; Rubin et al., 2007). The other CM2.0 chondrite in this study (LAP 02277) also exhibits strong particle alignment (Fig. 3). The histogram of the deviation of the MET 01070 particles’ long axes from the median azimuth shows a strong peak at ±10°; there are tall shoulders at ±20° and steeply declining tails at greater deviations on both sides of the peak (Fig. 5a). LAP 02277 is even more tightly peaked (Fig. 5b). Such high kurtosis or “peakedness” in the CM2.0 chondrites is indicative of a pronounced petrofabric. CM2.1 QUE 93005 has a very sharp peak at 10° deviation from the median azimuth and relatively broad flat tails flanking the peak (Fig. 5c). Less-altered CM chondrites have broader distributions, with decreasingly pronounced peaks flanking 0° deviation from the median azimuth (e.g., Fig. 5d–f). The least-altered CM chondrite in the study (CM2.6 QUE 97990; Fig. 1a) appears somewhat bimodal (Fig. 5f), although Murchison (CM2.5) (which is only slightly more altered) is broadly unimodal (Fig. 5e). Although neither QUE 97990 nor Murchison have noticeable petrofabrics, detailed analyses of Murchison have revealed a weak foliation (Fujimura et al., 1983; Hanna et al., 2012). Table 1 shows that the percentage of particles with long axes within 10° of the median azimuth tends to decrease with increasing petrologic subtype among the CM chondrites: i.e., the less-altered rocks tend to have fewer particles aligned in the same direction. Fig. 6 shows that particle alignment and CM subtype are strongly anticorrelated (r = 0.92, n = 9, 2a = 0.00014); this is significant at the 99.986% confidence level. It is clear that, with increasing degrees of aqueous alteration, CM chondrites exhibit more-pronounced petrofabrics. Additional support for this conclusion is found in Lindgren et al. (2012) who described CM2.0 SCO 06043 as having an obvious petrofabric caused by flattened chondrules that are elongated in a common direction.

185

MCY 05219 (CV3OxB), NWA 7107 (CV3R) and Y 981208 (CV3R) also have strong preferential orientations of their constituent particles (e.g., Fig. 4), similar to that of CM2.0 MET 01070 (Table 1). 3.6. Grain density in CM chondrites Grain density is defined as the mass of solid grains in a rock divided by the volume of solid grains (exclusive of pores); it is commonly expressed in units of g cm 3. Thus, meteorites with modally abundant low-density minerals (including phyllosilicates) will have a lower grain density than unaltered chondrites with few low-density minerals. It is therefore expected that grain density would correlate with the degree of alteration of CM chondrites because the more-altered samples should have higher modal abundances of phyllosilicates and other low-density materials (e.g., “PCP”). The least-altered chondrites should have higher grain densities. Using helium pycnometry, Macke et al. (2011) measured the grain densities of four of the CM chondrites included in this study. Although there are few data, Fig. 7 shows that grain density is indeed strongly correlated with petrologic subtype (r = 0.99, n = 4, 2a = 0.006); this correlation is significant at the 99.4% confidence level. 4. DISCUSSION 4.1. Collisional formation of petrofabrics in chondrites It is likely that, in general, chondrite petrofabrics are produced by impacts (Martin and Mills, 1980; Hamano and Yomogida, 1982; Sugiura and Strangway, 1983; Nakamura et al., 1992; Gattacceca et al., 2005; Rubin and Swindle, 2011). Among ordinary chondrites (OC) many rocks exhibit petrofabrics, commonly manifested as a strong foliation (a local planar structure) and a weaker lineation (a local linear structure) (Dodd, 1965; Sneyd et al., 1988). There are correlations in these OC between the strain ellipsoids (determined by the deformation of initially spheroidal chondrules) and magnetic-susceptibility ellipsoids (reflecting the alignment of metallic Fe–Ni grains); both parameters also correlate with shock indicators such as the degree of olivine deformation and the extent of rare-gas depletion (e.g., Sneyd et al., 1988; Gattacceca et al., 2005). CV3 Leoville has a prominent petrofabric as well as 0.1–2-lm-size matrix olivines with high densities of micro-cracks and dislocations (with Burgers vector b = [001]) (Nakamura et al., 1992). The conclusion that collisions are responsible for the creation of chondrite petrofabrics is supported by shock recovery experiments (up to 21 GPa peak shock pressures) on CV3OxA Allende and CM2.5 Murchison that produced whole-rock foliations and caused flattening of initially spheroidal chondrules (Nakamura et al., 1995; Tomeoka et al., 1999). Many of the olivine grains developed undulose extinction and planar fractures characteristic of shock-stage S3 (Sto¨ffler et al., 1991). As shown by Cain et al. (1986), chondrule flattening in a carbonaceous chondrite is

186

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

Fig. 5. Histograms of a suite of selected CM chondrites of different petrologic subtypes showing the deviation of particles (chondrules, chondrule fragments, chondrule pseudomorphs, refractory inclusions, amoeboid olivine inclusions, opaque grains, and “PCP” particles and clumps) from the median azimuth of each meteorite. (a) MET 01070 (CM2.0), one of the most altered CM chondrites, has a very sharp peak near the median azimuth, indicative of a pronounced petrofabric. (b) LAP 02277 (CM2.0) is also one of the most altered CM chondrites; it also has a prominent peak indicating a strong petrofabric. (c) QUE 93005 (CM2.1) is very altered, but has retained a few mafic silicate grains. It has a prominent peak near the median azimuth and a significant tail. The distribution appears slightly bimodal. (d) QUE 99355 (CM2.3) has experienced an intermediate amount of alteration. It has a broader peak than the more-altered samples. (e) Murchison (CM2.5) has experienced lesser degrees of alteration than the samples of lower petrologic subtype. There is a very broad unimodal peak indicating that the rock essentially lacks a petrofabric. (f) QUE 97990 (CM2.6), the least-altered CM chondrite in the study, has a broad bimodal distribution and shows no evidence of a whole-rock petrofabric.

associated with the presence of a whole-rock petrofabric. The proximate cause of chondrule flattening and whole-rock foliation in these rocks was probably the collapse of matrix pores under shock pressure and the squeezing of chondrules into pore spaces (Scott et al., 1992; Nakamura et al., 1995). (Porosity reduction during hypervelocity impacts is generally irreversible upon pressure release, e.g., Sharp and DeCarli, 2006.) Shock-recovery experiments show that chondrules are elongated in the plane of the shock front, roughly perpendicular to the compaction axis. Although not all CV3 chondrites have pronounced petrofabrics, those that do include reduced CV3R rocks

[Leoville (S3), Vigarano (S1–S2 breccia), Efremovka (S4), Arch (S3), NWA 7107 (S4) and Y 981208 (S4)] and oxidized rocks of the Bali subgroup (CV3OxB) [ALH 85006 (S3), Bali (S3), Grosnaja (S3), and MCY 05219 (S3)] (Martin et al., 1975; Cain et al., 1986; Scott et al., 1992; this study). Vigarano, which has the lowest shock stage among those CV chondrites with a petrofabric (Scott et al., 1992), was described by Martin et al. (1975) as having a lower degree of chondrule orientation than Leoville or Grosnaja. Because the shock stages of those CV chondrites studied by Scott et al. (1992) that do not have a petrofabric are all S1 (with the sole exception of ALH 81003, S2), it is clear

Fig. 6. Strong anticorrelation between the percentage of particles that have their long axes within 10° of the median azimuth versus subtype for nine CM chondrites. The diagram shows that the morealtered CM chondrites (i.e., those of lower petrologic subtype) have the greatest degree of particle alignment and, thus, the most prominent petrofabrics.

2.98 2.96

grain density

2.92

CV3R S4 94 60% 86% 94% 97% CM2.3 S1 297 38% 62% 80% 85%

QUE 99355 Cold Bokkeveld

CM2.2 S1 80 35% 64% 71% 71%

QUE 93005

CM2.1 S1 200 32% 52% 64% 71% CM2.0 ? 84 62% 87% 98% 98%

LAP 02277 MET 01070

CM2.0 ? 101 52% 81% 95% 96%

Murray

2.9 2.88 2.86 2.84 2.82 2.78

Classification Shock stage No. particles % Within 10° % Within 20° % Within 40° % Within 50°

Murchison

2.94

2.8

Table 1 Deviation of the long axis of individual particles from the median azimuth.

187

CV3R S4 48 54% 77% 98% 98%

NWA 7107

CM2.6 S1 450 10% 20% 48% 60%

CV3OxB S3 41 56% 80% 95% 95%

QUE 97990

CM2.5 S1 425 17% 36% 71% 83% CM2.4/2.5 S1 346 18% 33% 57% 69% CM2.4 S1 148 30% 56% 91% 96%

Y 791198

Murray

Murchison

MCY 05219

Y 981208

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

2.76

Nogoya Cold Bokkeveld

2.2

2.3

2.4

2.5

CM chondrite subtype

Fig. 7. Diagram of grain density (data from Macke et al. (2011)) versus subtype for the four CM chondrites that were common to the datasets in this paper and in Macke et al. (2011). There is a strong correlation, indicating that the least-altered CM chondrites (i.e., those with the highest subtypes) have the highest grain densities. This is presumably due to their low modal abundances of such low-density materials as phyllosilicates and “PCP”. The chondrites of high grain density have lesser degrees of hydration.

that there is a correlation between shock stage and the strength of the petrofabric among CV chondrites. 4.1.1. CM chondrites The two CM2.0 chondrites studied here do not contain mafic silicate grains and there is no currently accepted scale that can be used to measure the shock stages of these samples. If we exclude the two CM2.0 chondrites as being unmeasurable and accept my reclassification (as well as that of Tomeoka et al., 1999) of Murchison as S1, then 28 out of the 29 total “measurable” CM chondrites examined in this study and by Scott et al. (1992) are shock-stage S1; the sole remaining sample (EET 83250) is S2. Because CM chondrites are unrecrystallized and unequilibrated, their low shock stages cannot have resulted from significant postshock annealing of previously moderately to highly shocked

188

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

materials (as seems to be the case for many equilibrated S1 and S2 OC; Rubin, 2004). Therefore, it is clear that the CM chondrites in our collections never experienced high shock pressures; the pressures that they did experience were probably 5 GPa (Sto¨ffler et al., 1991). Because CM chondrites were only negligibly to lightly shocked, it seems possible that their petrofabrics were produced by processes other than impacts (e.g., Fujimura et al., 1983). Stacey et al. (1961) suggested that chondrite petrofabrics could have formed through uniaxial compaction caused by overburden lithostatic pressure. However, that process could not account for multiple episodes of deformation recorded in calcite grains in CM2.2-2.3 LON 94101 (Lindgren et al., 2010, 2011). Multiple collisions at shallow crustal levels were deemed more likely by Lindgren et al. (2011). Collisions have indeed affected CM chondrites; many of these rocks are breccias. Some (e.g., Murray, Cold Bokkeveld) contain multi-millimeter-size clasts exhibiting different degrees of aqueous alteration (Rubin and Wasson, 1986; Metzler et al., 1992). Some CM chondrites (e.g., QUE 99355, QUE 97990) contain detached clasts of dark mantle material that were probably broken off of chondrules (Fig. 11 of Trigo-Rodrı´guez et al. (2006)). Many CM chondrites (e.g., Cold Bokkeveld, Murchison, Murray, Nogoya, Pollen) are regolith breccias containing solar-windimplanted rare gases (Schultz and Kruse, 1989); others (e.g., Haripura, Mighei, Y 74662) are fragmental breccias (Bischoff and Sto¨ffler, 1992) that contain only “planetarytype” rare gases. Strictly elastic waves resulting from subsonic impacts can be ruled out as the cause responsible for the development of CM petrofabrics because such waves do not result in permanent deformation of the target (Melosh, 1989). It is more likely that shear resulting from hypervelocity impact events are responsible for creating the petrofabrics in CM chondrites (e.g., Gattacceca et al., 2005). However, the shock-stage S1 olivine grains in most CM chondrites (including those with significant petrofabrics) indicates that the shock recovery experiments of Tomeoka et al. (1999) might not be strictly applicable to real asteroids. The Murchison charge in these experiments underwent localized melting and its olivines reached the S3 level after the rock was shocked to 21 GPa. However, these experiments were necessarily characterized by narrow shock fronts; large colliding bodies would have much wider shock fronts. These could plausibly produce petrofabrics at lower shock pressures. It is possible that the shock wave had largely dissipated before it could deform the olivine crystal lattices. This is because CM chondrites are porous rocks: Krot et al. (2005) estimated the matrix abundance to be 70 vol.% matrix; Howard et al. (2009) determined the average total phyllosilicate abundance to be 75%. (The latter value is bound to be somewhat greater than the actual matrix abundance as it includes phyllosilicates within chondrule interiors and rims.) In porous bodies, the collisional kinetic energy is distributed through relatively small volumes of material (Stewart and Ahrens, 1999) and efficiently converted into heat in the crater vicinity (Melosh, 1989; Housen and

Holsapple, 1999). This is due to the high impedance mismatch between solids and voids (Kieffer, 1971). The increased PdV work done upon porous material during compression produces waste heat that is heterogeneously distributed (e.g., Melosh, 1989; Sharp and DeCarli, 2006). One potential consequence of heating an unequilibrated chondrite is that low-Ca clinopyroxene phenocrysts in chondrules could transform into orthopyroxene at low pressures if temperatures reached or exceeded 630 °C (Boyd and England, 1965; Grover, 1972). (Low-Ca clinopyroxene grains are petrographically recognizable because of their generally inclined extinction and the presence in them of polysynthetic twins parallel to (100); orthopyroxene grains have parallel extinction and lack polysynthetic twins.) Given the correlation in CM chondrites between the strength of the petrofabric and the subtype (Fig. 6), if petrofabric development was caused by the shock-induced collapsing of matrix pores, it is possible that the rocks with strong petrofabrics were heated above the orthopyroxene inversion temperature. If this were the case, there might be a correlation between the CM subtype and the percentage of pyroxene-bearing chondrules that contain low-Ca clinopyroxene. There is such a correlation (Table 2): (r = 0.97, n = 5, 2a = 0.003); it is significant at the 99.7% confidence level. The more-altered CM chondrites (those with the strongest petrofabrics) lack low-Ca clinopyroxene. Away from the impact site, shock waves attenuate rapidly. It seems likely that attenuated shock waves on the CM asteroid could cause the collapse of matrix pores and the squeezing of chondrules into pore spaces in matrix-rich samples (Scott et al., 1992; Nakamura et al., 1995) without causing most of the olivine grains to develop undulose extinction. Nevertheless, the waste heat generated by this process caused low-Ca clinopyroxene grains in the more collisionally affected target rocks to transform into orthopyroxene. The heating need not have been of long duration – Boyd and England (1965) routinely achieved clinoenstatite–orthoenstatite inversion in experiments lasting less than a week. If the CM chondrites had experienced sustained periods with temperatures P630 °C (i.e., within the 600– 700 °C range of metamorphic temperatures of type-4 ordinary chondrites; Table 4.4 of Dodd (1981)), they would likely have suffered phyllosilicate dehydration (e.g., Lipschutz et al., 1999), the decomposition of tochilinite into troilite (Fuchs et al., 1973) and graphitization of carbonaceous macromolecular matter (Kitajima et al., 2002). Because such transformations are in part dependent on the duration of heating (Kitajima et al., 2002), the paucity or absence of such effects indicates that temperature excursions in CM chondrites were brief. 4.1.2. CV chondrites The situation is different for CV chondrites. Because nearly every CV3 with a petrofabric is S3 or S4 and nearly every CV3 without a petrofabric is S1, it is obvious that hypervelocity impacts are responsible for the formation of petrofabrics in CV3 chondrites. I suggest that the lower modal abundance of matrix in CV3 chondrites (35 vol.%; McSween, 1977) compared to CM chondrites (70 vol.%; Krot et al., 2005) attenuates shock waves less

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

189

Table 2 CM chondrites and low-Ca clinopyroxene.

QUE 93005 Cold Bokkeveld Nogoya Murchison QUE 97990

Subtype

Pyroxene-bearing chondrules containing polysynthetically twinned low-Ca clinopyroxene

CM2.1 CM2.2 CM2.2 CM2.5 CM2.6

0/2 = 0% 0/4 = 0% 0/5 = 0% 11/20 = 55% 31/33 = 94%

effectively, allowing a greater volume of material in the crater vicinity to experience high shock pressures and for mafic silicates in these rocks to exhibit shock effects. CV chondrites still have sufficient amounts of matrix to allow pores to collapse under shock pressure and for chondrules to squeeze into pore spaces during shock events (Scott et al., 1992; Nakamura et al., 1995). These features are responsible for the correlation between shock stage and the possession of petrofabrics in CV3 chondrites. Nevertheless, my observations indicate that CV chondrites of all three subgroups contain polysynthetically twinned low-Ca clinoenstatite, indicating that, even in the shocked samples, matrix temperatures did not generally reach 630 °C; low-Ca clinopyroxene was preserved. 4.2. Relative timing of aqueous alteration and petrofabric formation There is petrographic evidence that bears on the question of whether CM chondrites were aqueously altered before or after the impact events that produced their petrofabrics. The Nogoya CM2.2 chondrite is an observed fall that has experienced little oxidation in the terrestrial environment. Many of the olivine grains in Nogoya are transected by irregular fractures, on the sides of which the olivine has been aqueously altered (Fig. 8). Most regions of individual olivine grains are fracture free and show no signs of alteration. It is apparent that the fractures in the olivine grains formed before the rock was altered. This

indicates that some impact events preceded aqueous alteration on the CM parent body. However, it cannot be established if the impact events that caused the olivine grains in Nogoya to fracture are the same collisions that were responsible for the production of petrofabrics in the CM chondrites. (Some chondrites experienced several distinct impact events; e.g., Lambert et al., 1984; Lindgren et al., 2011.) Stronger support for aqueous alteration having occurred after impact-fracturing can be found in the highly altered CM2.0 chondrite MET 01070. This meteorite contains several lenses up to 12 mm in length that range in thickness from 60–1000 lm (Fig. 1b; Rubin et al., 2007; Lindgren et al., 2012). The lenses are rich in PCP and associated Ni-bearing sulfide and contain some grains of Ca phosphate. They are subparallel to elongated fractures in the rock (e.g., Fig. 1b); their long axes point in approximately the same direction as the elongated chondrules. The lenses probably formed from anhydrous matrix material due to fluid flow along preexisting fractures. Although Scott et al. (1992) argued that energetic impacts into chondrites that already contained hydrous minerals would cause explosive expansion of these rocks upon the release of shock pressure, it is not clear that this process would have affected CM chondrites. The shock metamorphism experiments on CM2 Murchison by Tomeoka et al. (1999) did not result in the destruction of the Murchison charges. 4.3. Facilitation of aqueous alteration by impacts

Fig. 8. Isolated, moderately fractured olivine grain from Nogoya (CM2.2) showing that narrow bands of altered olivine (medium gray) flank the fractures (dark gray to black straight to curvy lines); the unfractured interior of the grain is unaltered. This implies that the fracturing preceded the aqueous alteration of the olivine.

4.3.1. CM chondrites I assume here that the textural evidence cited above for MET 01070 and Nogoya is dispositive and that, in general, the impact events that produced CM petrofabrics took place prior to aqueous alteration. If this is indeed the case, then the elongated phyllosilicates in CM chondrites described by Fujimura et al. (1982) probably formed within fractures in the whole rock created by impact events that occurred before alteration. Alternatively, the phyllosilicates may have been anhydrous, mafic-silicate-rich materials before the petrofabric-producing collisions that were altered only after subsequent sustained contact with aqueous fluids. Small-scale hypervelocity collisions occurred randomly at the surface of the CM parent asteroid and, although the shock waves attenuated rapidly away from the impact site, they caused some CM target regions to become more deformed and fractured than others. The more-deformed

190

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

regions developed stronger petrofabrics and more extensive fractures than regions that experienced less shear resulting from the impact events. They were also briefly heated to higher temperatures, allowing many of their low-Ca clinopyroxene grains to transform into orthopyroxene. Rock-physics experiments demonstrate significant increases in porosity (irreversible dilatancy) in silicate materials for a large range of stress states (e.g., Cristescu, 1989) due to the formation of fractures. Some CM chondrites were more severely affected by collisions than others. Shock pressures were generally too low to damage the olivine crystal lattices; these grains retained the sharp optical extinction characteristic of shock-stage S1. Water was eventually mobilized on the CM parent body, perhaps by subsequent impact-induced dehydration of phyllosilicates or by impact melting of ice (e.g., Lange et al., 1985). Some phyllosilicates may even have been produced during the impact (Furukawa et al., 2011). Bland et al. (2012) demonstrated that even low-energy impacts are capable of raising the temperature of chondrite matrix materials (in which phyllosilicates or ice could reside) several hundred degrees. The mobilized water seeped into essentially all of the CM chondrites, but more water was retained in those rocks that had more fractures, i.e., the particular CM chondrites that had been more significantly affected by impacts and exhibited stronger petrofabrics. These are the same rocks that became more aqueously altered, thereby accounting for the strong anticorrelation in CM chondrites between petrologic subtype and the preferred orientation of particles (Fig. 6). Phyllosilicate formation can occur rapidly. Michalopoulos and Aller (1995) found that K–Fe–Mg clay minerals can form on solid substrates within 1–3 years at 28 °C. Many fresh volcanic rocks contain clay minerals lining fractures and mineral cleavages, within vesicles, on the walls of voids and as olivine pseudomorphs; these clays formed rapidly at more elevated temperatures (e.g., Ku¨hnel and Van der Gaast, 1989). It seems likely that the phyllosilicates in CM chondrites also formed relatively quickly from the aqueous fluids that seeped into pre-existing shock-induced fractures. The weak inverse correlation between D17O and CM petrologic subtype (Fig. 8 of Rubin et al., 2007) indicates that the water responsible for aqueous alteration was enriched in heavy O isotopes. This is consistent with the report of Airieau et al. (2005) that the D17O and d18O values of watersoluble salts in CM chondrites correlate with the extent of whole-rock alteration; water with heavy O isotopes exchanged oxygen with silicate phases. The stochastic nature of impact and alteration processes would suggest that different CM chondrites may have been altered at different times. This is consistent with the results of Tyra et al. (2009) who determined Mn–Cr systematics in carbonate grains from two, probably paired, CM2.0 chondrites (ALH 84051 and ALH 84034) and found that carbonate formation (and, hence, aqueous alteration; e.g., Kerridge and Bunch, 1979) occurred 5 Ma apart. Such differences in the timing of alteration are difficult to reconcile with global models of aqueous alteration (e.g., Grimm and McSween, 1989; Young et al., 1999; Cohen and Coker,

2000; McSween et al., 2002; Travis and Schubert, 2005), but are readily accounted for by discrete impact events. The present model implies that there should be fracturecontrolled fluid flow in the CM chondrites. There is petrographic evidence for this process: (1) The PCP-rich lenses in MET 01070 (Fig. 1b) are subparallel to most of the elongated fractures and elongated chondrules in the rock and probably formed through fluid flow. (2) Carbonate veins in the CM2.0 SCO 06043 (Lindgren et al., 2012), CM2.0 ALH84051 (Tyra et al., 2009) and CM2.2-2.3 LON 94101 (Lindgren et al., 2010, 2011) provide additional evidence of fluid flow, presumably through preexisting fractures (e.g., Lindgren et al., 2011). (3) Aligned aragonite crystals in CM2.4/2.5 Murray were formed during late-stage aqueous alteration from fluids that were “focused within zones of high porosity and permeability along a weak compactional fabric in the matrix” (Lee and Ellen, 2008). 4.3.2. CV chondrites The situation is somewhat different for CV chondrites. Among the CV chondrites with petrofabrics, roughly half are members of the reduced subgroup (CV3R) and roughly half are members of the Bali-like oxidized subgroup (CV3OxB); none is a member of the (less-altered) Allendelike oxidized subgroup (CV3OxA). CV chondrites contain much less water (bound as hydroxyl in phyllosilicate) than CM chondrites: the Allende and Bali falls contain <0.1 and 0.30 wt.% H2O+ (indigenous water), respectively; in contrast, the Murchison CM fall contains 9 wt.% H2O+ (Jarosewich, 1990). This is also reflected in the total phyllosilicate abundance: CV3R (0 vol.%), CV3OxA (1.9 vol.%), CV3OxB (3.8 vol.%), CM (75 vol.%) (Howard et al., 2010, 2011). I suggest that the reduced CV chondrites formed from material that had been previously compacted by impacts. The porosities of CV3R samples are 0.6–8% (Macke et al., 2011). Hypervelocity impacts into this compacted material shocked many of the reduced CV chondrites to S3–S4 levels and produced petrofabrics within them as the relatively small amounts of pores collapsed and chondrules were squeezed into the available pore spaces. The CV3R chondrites remained relatively unaltered and unoxidized even after water was subsequently mobilized by impact heating of phyllosilicates or ice; this is because the low porosities of these rocks permitted relatively little water to seep into them to facilitate alteration. In contrast, the members of the two oxidized CV subgroups formed from more porous, less-compacted materials – their porosities are typically 20–28% (Macke et al., 2011). Some of these more-porous CV chondrites (i.e., the CV3OxB samples) were significantly shocked, typically reaching shock-stage S3. They were heavily fractured and developed strong petrofabrics. When water was subsequently mobilized, they became greatly altered because there were many sites where water could be retained. Abundant phyllosilicates formed in the matrix (Tomeoka and Buseck, 1982, 1990; Keller and Buseck, 1990; Zolensky et al., 1993) and most of the metal was replaced by magnetite. The slight depletion in 16O of the CV3OxB samples relative to the O-isotopic composition of the other subgroups

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

(Weisberg et al., 1997) reflects more-complete isotopic exchange in the Bali subgroup between initially unaltered CM materials and 16O-poor water. The Allende-like CV3OxA samples formed from similar porous, uncompacted materials but did not experience appreciable shock. They remained essentially unshocked (shock-stage S1). They were not extensively fractured and did not develop noticeable petrofabrics. Nevertheless, because of their high initial porosity, the CV3OxA samples became significantly altered when water was later mobilized. Secondary phases such as hedenbergite and andradite were formed in porous CV materials at this time (e.g., within Allende and some regions of Vigarano; MacPherson and Krot, 2002). Because the CV3OxA samples were not as fractured (by shock) as the CV3OxB samples, the overall degree of alteration in CV3OxA rocks was somewhat less. Phyllosilicates formed much less abundantly in the CV3OxA matrix regions and less metallic Fe–Ni was converted into magnetite. It is not clear what is responsible for the differences in porosity between the reduced CV subgroup and the two oxidized CV subgroups, but there are analogs among OC. Whereas the typical porosities of OC are 6–10% (Britt and Consolmagno, 2003), there are several OC with much higher porosities, e.g., Baszko´wka – 19%, NWA 2380 – 19% and Miller (Arkansas) – 20% (Friedrich et al., 2008; Sasso et al., 2009). High-porosity OC and CV chondrites may have retained some of their initial nebular porosity because they experienced less parent-body impact-compaction than the more common lower-porosity members of their groups. 5. CONCLUSIONS Nearly every CM chondrite is shock-stage S1, including many of the samples that exhibit significant particle alignments. There is a consensus that petrofabrics in chondrites are produced by collisions; thus, it seems likely that the high porosity of the CM asteroid (as reflected by the high modal abundance of matrix: 70 vol.%) caused sufficient attenuation of the shock waves from a neighboring impact site to avoid damaging olivine crystal lattices. The shock waves were still energetic enough to produce whole-rock petrofabrics by pore collapse and the squeezing of chondrules into pore spaces. Temperature increases associated with the impacts caused clinopyroxene to transform into orthopyroxene in the more-strongly shocked rocks. Randomly spaced collisions on the CM parent asteroid created fractures and produced petrofabrics in some near-surface regions. Subsequent mobilization of water (caused by impact-induced dehydration of existing phyllosilicates or by the impact melting of ice) caused seepage into CM-chondrite materials. Water may have been preferentially retained in the more-fractured regions of the near-surface environment (i.e., in those regions most affected by impacts and that had the most pronounced petrofabrics). The water would have caused alteration of pre-existing materials, mainly mafic silicates, metal grains and chondrule glass. The less-deformed, less-fractured CM regions (i.e., those regions that were less affected by impacts) experienced lower degrees of aqueous alteration. This produced a strong anti-

191

correlation between petrologic subtype and the degree of particle alignment in CM whole rocks. The CV3R chondrites formed from low-porosity material that had been previously compacted by impacts on the parent asteroid. These rocks were typically shocked to S3-S4 levels and developed petrofabrics. After water was mobilized, there were relatively few pore spaces in the CV3R samples to accommodate the water; the CV3R chondrites thus remained relatively unaltered and unoxidized. The two oxidized CV subgroups formed from higherporosity materials. Some of these rocks (i.e., the CV3OxB samples) were significantly shocked, typically reaching shock-stage S3. They were appreciably fractured and developed petrofabrics. When water was subsequently mobilized, they became greatly altered because there were many sites where water could be retained. Other regions of porous CV material (i.e., the CV3OxA samples) escaped significant shock; they were not extensively fractured and did not develop petrofabrics. Their high initial porosity facilitated oxidation during aqueous alteration, but because the CV3OxA samples were not as fractured as the CV3OxB samples, they were altered to a lesser extent. ACKNOWLEDGMENTS I am grateful to the curators at the Smithsonian Institution, the NASA-Johnson Space Center, NIPR in Japan, and A. Greshake of the Museum fu¨r Naturkunde in Berlin for the loan of thin sections. I also thank H. J. Melosh and J. T. Wasson for useful comments and K. T. Howard and an anonymous referee for their helpful reviews. This work was supported by NASA Cosmochemistry Grant NNG06GF95G.

REFERENCES Airieau S. A., Farquhar J., Thiemens M. H., Leshin L. A., Bao H. and Young E. (2005) Planetesimal sulfate and aqueous alteration in CM and CI carbonaceous chondrites. Geochim. Cosmochim. Acta 69, 4166–4171. Baker L., Franchi I. A., Wright I. P. and Pillinger C. T. (2002) The oxygen isotopic composition of water from Tagish Lake: its relationship to low-temperature phases and to other carbonaceous chondrites. Meteorit. Planet. Sci. 37, 977–985. Barber D. J. (1981) Matrix phyllosilicates and associated minerals in C2M carbonaceous chondrites. Geochim. Cosmochim. Acta 45, 945–970. Barber D. J. (1985) Phyllosilicates and other layer-structured materials in stony meteorites. Clay Miner. 20, 415–454. Beck P., Quirico E., Montes-Hernandez G., Bonal L., Bollard J., Orthous-Daunay F.-R., Howard K. T., Schmitt B., Brissaud O., Deschamps F., Wunder B. and Guillot S. (2010) Hydrous mineralogy of CM and CI chondrites from infrared spectroscopy and their relationship with low albedo asteroids. Geochim. Cosmochim. Acta 74, 4881–4892. Benedix G. K., Leshin L. A., Farquhar J., Jackson T. and Thiemens M. H. (2003) Carbonates in CM2 chondrites: constraints on alteration conditions from oxygen isotopic compositions and petrographic observations. Geochim. Cosmochim. Acta 67, 1577–1588. Bischoff A. and Sto¨ffler D. (1992) Shock metamorphism as a fundamental process in the evolution of planetary bodies: information from meteorites. Eur. J. Mineral. 4, 707–755.

192

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

Bischoff A., Scott E. R. D., Metzler K. and Goodrich C. A. (2006) Nature and origins of meteoritic breccias. In Meteorites and the Early Solar System II (eds. D. S. Lauretta and H. Y. McSween). University of Arizona Press, Tucson, pp. 679–712. Bland P. A., Jackson M. D., Coker R. F., Cohen B. A., Webber J. B. W., Lee M. R., Duffy C. M., Chater R. J., Ardakani M. G., McPhail D. S., McComb D. W. and Benedix G. K. (2009) Why aqueous alteration in asteroids was isochemical: high porosity – high permeability. Earth Planet. Sci. Lett. 287, 559–568. Bland P. A., Muxworthy A. R., Collins G. S., Moore J., Davison T. M., Prior D. J., Wheeler J., Ciesla F. J. and Dyl K. A. (2012) Effect of low intensity impacts on chondrite matrix. Lunar Planet. Sci. 43, #2005 (abstr.). Bourot-Denise M., Zanda B., Marrocchi Y., Greenwood R. C., Pont S., Hewins R. H., Franchi I. A. and Cornen G. (2010) Paris: The slightly altered, slightly metamorphosed CM that bridges the gap between CMs and COs. Lunar Planet. Sci. 41, #1683 (abstr.). Boyd F. R. and England J. L. (1965) The rhombic enstatiteclinoenstatite inversion. Ann. Rpt. Dir. Geophys. Lab. Washington DC, Carnegie Institution of Washington Yrbk 64, 1964– 1965, pp. 117–120. Britt D. T. and Consolmagno G. J. (2003) Stony meteorite porosities and densities: a review of the data through 2001. Meteorit. Planet. Sci. 38, 1161–1180. Browning L., McSween H. and Zolensky M. (1996) Correlated alteration effects in CM carbonaceous chondrites. Geochim. Cosmochim. Acta 60, 2621–2633. Browning L., McSween H. Y. and Zolensky M. E. (2000) On the origin of rim textures surrounding anhydrous silicate grains in CM carbonaceous chondrites. Meteorit. Planet. Sci. 35, 1015– 1023. Bunch T. and Chang S. (1980) Carbonaceous chondrites – II: carbonaceous chondrite phyllosilicates and light element geochemistry as indicators of parent body processes and surface conditions. Geochim. Cosmochim. Acta 44, 1543–1577. Cain P., McSween H. Y. and Woodward N. B. (1986) Structural deformation of the Leoville chondrite. Earth Planet. Sci. Lett. 77, 165–175. Ciesla F. and Lauretta D. (2005) Radial migration and dehydration of phyllosilicates in the solar nebula. Earth Planet. Sci. Lett. 231, 1–8. Clayton R. N. and Mayeda T. K. (1984) The oxygen isotope record in Murchison and other carbonaceous chondrites. Earth Planet. Sci. Lett. 67, 151–161. Clayton R. N. and Mayeda T. K. (1999) Oxygen isotope studies of carbonaceous chondrites. Geochim. Cosmochim. Acta 63, 2089– 2104. Cohen B. A. and Coker R. F. (2000) Modeling of liquid water on CM meteorite parent bodies and implications for amino acid racemization. Icarus 145, 369–381. Coker R. F. and Cohen B. A. (2001) The effect of liquid transport on the modeling of CM parent bodies (abstract). Meteorit. Planet. Sci. 36, A443–A444. Cristescu N. (1989) Rock Rheology. Kluwer Academic Publishers, Dordrecht, Holland, 336 pp. Dodd R. T. (1965) Preferred orientation of chondrules in chondrites. Icarus 4, 308–316. Dodd R. T. (1981) Meteorites: A Petrologic-Chemical Synthesis. Cambridge University Press, Cambridge, 368 pp. Friedrich J. M., Macke R. J., Wignarajah D. P., Rivers M. L., Britt D. T. and Ebel D. S. (2008) Pore size distribution in an uncompacted equilibrated ordinary chondrite. Planet. Space Sci. 56, 895–900.

Fuchs L. H., Olsen E. and Jensen K. J. (1973) Mineralogy, mineral chemistry, and composition of the Murchison (C2) meteorite. Smithson. Contrib. Earth Sci. 10, 1–39. Fujimura A., Kato M. and Kumazawa M. (1982) Preferred orientation of phyllosilicates in Yamato-74642 and -74662, in relation to deformation of C2 chondrites. Proc. Symp. Ant. Met. 7, 207–215. Fujimura A., Kato M. and Kumazawa M. (1983) Preferred orientation of phyllosilicate [001] in matrix of Murchison meteorite and possible mechanisms of generating the oriented texture in chondrites. Earth Planet. Sci. Lett. 66, 25–32. Furukawa Y., Sekine T., Kakegawa T. and Nakazawa H. (2011) Impact-induced phyllosilicate formation from olivine and water. Geochim. Cosmochim. Acta 75, 6461–6472. Gattacceca J., Rochette P., Denise M., Consolmagno G. and Folco L. (2005) An impact origin for the foliation of chondrites. Earth Planet. Sci. Lett. 234, 351–368. Greenwood R. C., Lee M. R., Hutchison R. and Barber D. J. (1994) Formation and alteration of CAIs in Cold Bokkeveld (CM2). Geochim. Cosmochim. Acta 58, 1913–1935. Grimm R. E. and McSween H. Y. (1989) Water and the thermal evolution of carbonaceous chondrite parent bodies. Icarus 82, 244–280. Grover J. E. (1972) The stability of low-clinoenstatite in the system Mg2Si2O6–CaMgSi2O6 (abstract). Trans. Am. Geophys. Union 53, 539. Guo W. and Eiler J. M. (2007) Temperatures of aqueous alteration and evidence for methane generation on the parent bodies of the CM chondrites. Geochim. Cosmochim. Acta 71, 5565–5575. Hamano Y. and Yomogida K. (1982) Magnetic anisotropy and porosity of Antarctic chondrites. Proc. Symp. Ant. Met. 7, 281– 290. Hanna R. D., Ketcham R. A. and Hamilton V. E. (2012) Inclusion foliation in Murchison as revealed by high resolution x-ray CT. Lunar Planet. Sci. 43, #1242 (abstr.). Hanowski N. P. and Brearley A. J. (2001) Aqueous alteration of chondrules in the CM carbonaceous chondrite, Allan Hills 81002: implications for parent body alteration. Geochim. Cosmochim. Acta 65, 495–518. Housen K. R. and Holsapple K. A. (1999) Impact cratering on porous low-density bodies. Lunar Planet. Sci. 30, 1228. Howard K. T., Benedix G. K., Bland P. A. and Cressey G. (2009) Modal mineralogy of CM2 chondrites by X-ray diffraction (PSD-XRD). Part 1: total phyllosilicate abundance and the degree of aqueous alteration. Geochim. Cosmochim. Acta 73, 4576–4589. Howard K. T., Benedix G. K., Bland P. A. and Cressey G. (2010) Modal mineralogy of CV3 chondrites by X-ray diffraction (PSD-XRD). Geochim. Cosmochim. Acta 74, 5084–5097. Howard K. T., Benedix G. K., Bland P. A. and Cressey G. (2011) Modal mineralogy of CM chondrites by X-ray diffraction (PSD-XRD): part 2. Degree, nature and settings of aqueous alteration. Geochim. Cosmochim. Acta 75, 2735–2751. Jarosewich E. (1990) Chemical analyses of meteorites – a compilation of stony and iron meteorite analyses. Meteoritics 25, 323–337. Keller L. P. and Buseck P. R. (1990) Aqueous alteration in the Kaba CV3 carbonaceous chondrite. Geochim. Cosmochim. Acta 54, 2113–2120. Kerridge J. F. and Bunch T. E. (1979) Aqueous alteration on asteroids: evidence from carbonaceous meteorites. In Asteroids (ed. T. Gehrels). Univ. Arizona Press. Tueson, Arizona, pp. 745-764. Kieffer S. W. (1971) Shock metamorphism of the Coconino sandstone at Meteor Crater, Arizona. J. Geophys. Res. 76, 5449–5473.

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194 Kitajima F., Nakamura T., Takaoka N. and Murae T. (2002) Evaluating the thermal metamorphism of CM chondrites by using the pyrolytic behavior of carbonaceous macromolecular matter. Geochim. Cosmochim. Acta 66, 163–172. Kracher A., Keil K., Kallemeyn G. W., Wasson J. T., Clayton R. N. and Huss G. I. (1985) The Leoville (CV3) accretionary breccia. Lunar Planet. Sci. Conf. 16, D123–D135. Krot A. N. and Wasson J. T. (1995) Igneous rims on low-FeO and high-FeO chondrules in ordinary chondrites. Geochim. Cosmochim. Acta 59, 4951–4966. Krot A. N., Petaev M. I., Scott E. R. D., Choi B.-G., Zolensky M. E. and Keil K. (1998) Progressive alteration in CV3 chondrites: more evidence for asteroidal alteration. Meteorit. Planet. Sci. 33, 1065–1085. Krot A. N., Keil K., Goodrich C. A., Scott E. R. D. and Weisberg M. K. (2005) Classification of meteorites. In Meteorites, Comets, and Planets (ed. A. M. Davis). Elsevier, Amsterdam, pp. 83–129. Ku¨hnel R. A. and Van der Gaast S. J. (1989) Formation of clay minerals by mechanochemical reactions during grinding of basalt under water. Appl. Clay Sci. 4, 295–305. Lambert P., Lewis C. and Moore C. B. (1984) Repeated shock and thermal metamorphism of the Abernathy meteorite. Meteoritics 19, 29–48. Lange M. A., Lambert P. and Ahrens T. J. (1985) Shock effects on hydrous minerals and implications for carbonaceous meteorites. Geochim. Cosmochim. Acta 49, 1715–1726. Lee M. R. and Ellen R. (2008) Aragonite in the Murray (CM2) carbonaceous chondrite: implications for parent body compaction and aqueous alteration. Meteorit. Planet. Sci. 43, 1219– 1231. Lindgren P., Lee M. R., Sofe M. and Burchell M. (2010) Microstructures of a calcite vein in LON 94101: evidence for asteroidal fluid flow? Meteorit. Planet. Sci. 45, A119 (abstract). Lindgren P., Lee M. R., Sofe M. and Burchell M. J. (2011) Microstructure of calcite in the CM2 carbonaceous chondrite LON 94101: implications for deformation history during and/or after aqueous alteration. Earth Planet. Sci. Lett. 306, 289–298. Lindgren P., Lee M. R. and Sofe M. (2012) Evidence for multiple fluid pulses in the CM1 carbonaceous chondrite parent body. Lunar Planet. Sci. 43, #1949 (abstr.). Lipschutz M. E., Zolensky M. E. and Bell M. S. (1999) New petrographic and trace element data on thermally metamorphosed carbonaceous chondrites. Antarct. Meteorit. Res. 12, 57–80. Macke R. J., Consolmagno G. J. and Britt D. T. (2011) Density, porosity, and magnetic susceptibility of carbonaceous chondrites. Meteorit. Planet. Sci. 46, 1842–1862. MacPherson G. J. and Krot A. N. (2002) Distribution of Ca–Fe– silicates in CV3 chondrites: possible controls by parent-body compaction. Meteorit. Planet. Sci. 37, A91 (abstract). Martin P. M. and Mills A. A. (1980) Preferred chondrule orientations in meteorites. Earth Planet. Sci. Lett. 51, 18–25. Martin P. M., Mills A. A. and Walker E. (1975) Preferential orientation in four C3 chondritic meteorites. Nature 257, 37–38. McSween H. Y. (1977) Petrographic variations among carbonaceous chondrites of the Vigarano type. Geochim. Cosmochim. Acta 41, 1777–1790. McSween H. Y. (1979) Alteration in CM carbonaceous chondrites inferred from modal and chemical variations in matrix. Geochim. Cosmochim. Acta 43, 1761–1770. McSween H. Y., Ghosh A., Grimm R. E., Wilson L. and Young E. D. (2002) Thermal evolution models of asteroids. In: Asteroids III (ed. Bottke W. F., Cellino A., Paolicchi P. and Binzel R. P.). pp. 559–571. Univ. Arizona Press, Tucson.

193

Melosh H. J. (1989) Impact Cratering: A Geologic Process. Oxford University Press, New York, 245 pp. Metzler K. (2004) Formation of accretionary dust mantles in the solar nebula: Evidence from preirradiated olivines in CM chondrites. Geochim. Cosmochim. Acta 39, 1307–1319. Metzler K., Bischoff A. and Sto¨ffler D. (1992) Accretionary dust mantles in CM chondrites: evidence for solar nebula processes. Geochim. Cosmochim. Acta 56, 2873–2897. Michalopoulos P. and Aller R. C. (1995) Rapid clay mineral formation in Amazon Delta sediments: reverse weathering and oceanic elemental cycles. Science 270, 614–617. Mu¨ller W. F. and Wlotzka F. (1982) Mineralogical study of the Leoville meteorite (CV3): macroscopic texture and transmission electron microscopic observations (abstract). Lunar Planet. Sci. 13, 558–559. Nakamura T., Tomeoka K. and Takeda H. (1992) Shock effects of the Leoville CV carbonaceous chondrite: a transmission electron microscope study. Earth Planet. Sci. Lett. 114, 159–170. Nakamura T., Tomeoka K., Sekine T. and Takeda H. (1995) Impact-induced chondrule flattening in the Allende CV3 carbonaceous chondrite: shock experiments. Meteoritics 30, 344–347. Palmer E. E. and Lauretta D. S. (2011) Aqueous alteration of kamacite in CM chondrites. Meteorit. Planet. Sci. 46, 1587– 1607. Petaev M. I. and Wood J. A. (1998) The condensation with partial isolation (CWPI) model of condensation in the solar nebula. Meteorit. Planet. Sci. 33, 1123–1137. Rosenberg N. D., Browning L. and Bourcier W. L. (2001) Modeling aqueous alteration of CM carbonaceous chondrites. Meteorit. Planet. Sci. 36, 239–244. Rubin A. E. (1984) Coarse-grained chondrule rims in type 3 chondrites. Geochim. Cosmochim. Acta 48, 1779–1789. Rubin A. E. (2004) Post-shock annealing and post-annealing shock in equilibrated ordinary chondrites: implications for the thermal and shock histories of chondritic asteroids. Geochim. Cosmochim. Acta 68, 673–689. Rubin A. E. (2007) Petrography of refractory inclusions in CM2.6 QUE 97990 and the origin of melilite-free spinel inclusions in CM chondrites. Meteorit. Planet. Sci. 42, 1711–1726. Rubin A. E. (2010) Physical properties of chondrules in different chondrite groups: implications for multiple melting events in dusty environments. Geochim. Cosmochim. Acta 74, 4807–4828. Rubin A. E. and Swindle T. D. (2011) Flattened chondrules in the LAP 04581 LL5 chondrite: evidence for an oblique impact into LL3 material and subsequent collisional heating. Meteorit. Planet. Sci. 476, 587–600. Rubin A. E. and Wasson J. T. (1986) Chondrules in the Murray CM2 meteorites and compositional differences between CM– CO and ordinary chondrite chondrules. Geochim. Cosmochim. Acta 50, 307–315. Rubin A. E., Trigo-Rodrı´guez J. M., Huber H. and Wasson J. T. (2007) Progressive aqueous alteration of CM carbonaceous chondrites. Geochim. Cosmochim. Acta 71, 2361–2382. Sasso M. R., Macke R. J., Britt D. T., Rivers M. L., Ebel D. S. and Friedrich J. M. (2009) Physical properties of incompletely compacted equilibrated ordinary chondrites: implications for asteroidal structure and impact processing. Lunar Planet. Sci. 40, #1670 (abstr.). Scherer P. and Schultz L. (2000) Noble gas record, collisional history, and pairing of CV, CO., CK, and other carbonaceous chondrites. Meteorit. Planet. Sci. 35, 145–153. Schultz L. and Kruse H. (1989) Helium, neon, and argon in meteorites – a data compilation. Meteoritics 24, 155–172.

194

A.E. Rubin / Geochimica et Cosmochimica Acta 90 (2012) 181–194

Scott E. R. D., Keil K. and Sto¨ffler D. (1992) Shock metamorphism of carbonaceous chondrites. Geochim. Cosmochim. Acta 56, 4281–4293. Sharp T. G. and De Carli P. S. (2006) Shock effects in meteorites. In Meteorites and the Early Solar System II (eds. D. S. Lauretta and H. Y. McSween). University of Arizona Press, Tucson, pp. 653–677. Sneyd D. S., McSween H. Y., Sugiura N., Strangway D. W. and Nord G. L. (1988) Origin of petrofabrics and magnetic anisotropy in ordinary chondrites. Meteoritics 23, 139– 149. Stacey F. D., Lovering J. F. and Parry L. G. (1961) Thermomagnetic properties, natural magnetic moments, and magnetic anisotropies of some chondritic meteorites. J. Geophys. Res. 66, 1523–1534. Stewart S. T. and Ahrens T. J. (1999) Porosity effects on impact processes in solar system materials. Lunar Planet. Sci. 30, #2020 (abstr.). Sto¨ffler D., Keil K. and Scott E. R. D. (1991) Shock metamorphism of ordinary chondrites. Geochim. Cosmochim. Acta 55, 3845– 3867. Sugiura N. and Strangway D. W. (1983) Magnetic anisotropy and porosity of chondrites. Geophys. Res. Lett. 10, 83–86. Tomeoka K. and Buseck P. R. (1982) Intergrown mica and montmorillonite in the Allende carbonaceous chondrite. Nature 299, 326–327. Tomeoka K. and Buseck P. R. (1985) Indicators of aqueous alteration in CM carbonaceous chondrites: Microtextures of a layered mineral containing Fe, S, O and Ni. Geochim. Cosmochim. Acta 49, 2149–2163. Tomeoka K. and Buseck P. R. (1990) Phyllosilicates in the Mokoia CV carbonaceous chondrite: evidence for aqueous alteration in an oxidizing condition. Geochim. Cosmochim. Acta 54, 1745– 1754. Tomeoka K., Yamahana Y. and Sekine T. (1999) Experimental shock metamorphism of the Murchison CM carbonaceous chondrite. Geochim. Cosmochim. Acta 63, 3683–3703. Travis B. J. and Schubert G. (2005) Hydrothermal convection in carbonaceous chondrite parent bodies. Earth Planet. Sci. Lett. 240, 234–250.

Trigo-Rodrı´guez J. M., Rubin A. E. and Wasson J. T. (2006) Nonnebular origin of dark mantles around chondrules and inclusions in CM chondrites. Geochim. Cosmochim. Acta 70, 1271– 1290. Tyra M. A., Brearley A. J., Hutcheon I. D., Ramon E., Matzel J. and Weber P. (2009) Carbonate formation timescales vary between CM1 chondrites ALH84051 and ALH84034. Lunar Planet. Sci. 40, #2474 (abstr.). Wasson J. T. (2008) Evaporation of nebular fines during chondrule formation. Icarus 195, 895–907. Wasson J. T. and Trigo-Rodrı´guez J. M. (2004) Evaporation during chondrule formation, recondensation as fine particles, and the condensation of S and other volatile elements. Lunar Planet. Sci. 35, #2140 (abstr.). Weisberg M. K., Prinz M., Clayton R. N. and Mayeda T. K. (1997) CV3 chondrites: three subgroups, not two (abstract). Meteorit. Planet. Sci. 32, A138–A139. Young E. D. (2001) The hydrology of carbonaceous chondrite parent bodies and the evolution of planet progenitors. Phil. Trans. Roy. Soc. London A 359, 2095–2110. Young E. D., Ash R. D., England P. and Rumble D. R. (1999) Fluid flow in chondritic parent bodies: deciphering the compositions of planetesimals. Science 286, 1331–1335. Zanda B., Bourot-Denise M., Hewins R. H., Barrat J. A. and Gattacceca J. (2010) The Paris CM chondrite yields new insights on the onset of parent body alteration. 73rd Meeting of the Meteoritical Society. 5312(abstr.). Zolensky M. E., Barrett R. A. and Browning L. B. (1993) Mineralogy and composition of matrix and chondrule rims in carbonaceous chondrites. Geochim. Cosmochim. Acta 57, 3123– 3148. Zolensky M. E., Mittlefehldt D. W., Lipschutz M. E., Wang M.-S., Clayton R. N., Mayeda T. K., Grady M. M., Pillinger C. T. and Barber D. (1997) CM chondrites exhibit the complete petrologic range from type 2 to 1. Geochim. Cosmochim. Acta 61, 5099–5115. Associate editor: Anders Meibom