Alex M Ruzicka

Comparative geochemistry of basalts from the moon, earth, HED asteroid, and Mars: implications for the origin of the moon

Abstract Most hypotheses for the origin of the Moon (rotational fission, co-accretion, and collisional ejection from the Earth, including “giant impact”) call for the formation of the Moon in a geocentric environment. However, key geochemical data for basaltic rocks from the Moon, Earth, the howardite-eucrite-diogenite (HED) meteorite parent body (probably asteroid 4-Vesta), and the shergottite-nakhlite-chassignite (SNC) meteorite parent body (likely Mars), provide no evidence that the Moon was derived from the Earth, and suggest that some objects with lunar-like compositions were produced without involvement of the Earth. The source region compositions of basalts produced in the Moon (mare basalts) were similar to those produced in the HED asteroid (eucrites) with regard to volatile-lithophile elements (Na, K, Rb, Cs, and Tl), siderophile elements (Ni, Co, Ga, Ge, Re, and Ir), and ferromagnesian elements (Mg, Fe, Cr, and V), and less similar to those in the Earth or Mars. Mare and eucrite basalts differ in their Mn abundances, Fe/Mn values, and isotopic composition, suggesting that the Moon and HED asteroid formed in different nebular locations. However, previous claims that the Moon and HED parent body differ significantly in the abundances of some elements, such as Ni, Co, Cr, and V, are not supported by the data. Instead, Cr-Mg-Fe-Ni-Co abundance systematics suggest a close similarity between the source region compositions and conditions involved in producing mare and eucrite basalts, and a significant difference from those of terrestrial basalts. The data imply that the Moon and HED asteroid experienced similar volatile-element depletion and similar fractionation of metallic and mafic phases. Among hypotheses of lunar origin, rotational fission, and small-impact collisional ejection seem less tenable than co-accretion, capture, or a variant of giant-impact collisional ejection in which the Moon inherits the composition of the impactor. Both the Moon and HED asteroid may have been derived from a class of objects that were common in the early solar system. “The most plausible model for the origin of the Moon in line with geochemical and cosmochemical constraints is an impact-induced “fission” of the proto-Earth.” — Wanke and Dreibus (1986 ) “Clearly, the Moon and eucrite parent body resemble each other to a high degree. Nature produced such a composition not once but at least twice. This calls into question an entire class of models that invoke ad hoc processes to explain the Moon by a unique chance event.” — Anders (1977 ) “The similarity between eucrites and lunar mare basalts are remarkable. Were it not for the differences in age and oxygen-isotope signature, it might be difficult to distinguish them on petrological or geochemical grounds.” — Taylor (1986 )

Mega-chondrules and large, igneous-textured clasts in Julesberg (L3) and other ordinary chondrites: vapor-fractionation, shock-melting, and chondrule formation

Abstract A petrographic-microprobe study of large, metal-poor, igneous-textured objects in Julesberg (L3) and other ordinary chondrites suggests that they can be classified into two petrographic types: mega-chondrules and large lithic clasts; and into two chemical types: Na-poor and Na-rich. Mega-chondrules show textural evidence of having solidified as freely-floating melt droplets, whereas lithic clasts formed by the fragmentation of larger objects, possibly still larger mega-chondrules. Barred-olivine or barred-olivine-pyroxene textures are most common for mega-chondrules, whereas a variety of textures occur in large lithic clasts. The two petrographic types cannot be distinguished on the basis of modal, bulk, or phase compositions. Sodium-poor objects are characterized by (1) plagioclase or glass with mainly bytownite composition (An70-90 typical), (2) subchondritic Na/Al ratios and typically subchondritic volatile-element (Na, K, and Mn) abundances, and (3) bulk-chemical trends that resemble those expected for vapor-fractionation processes. Some Na-poor objects may have formed by the melting of precursors that formed as condensates or vaporization residues; others may have formed by the melting of precursors that formed by fractional condensation or fractional vaporization. Following vapor-fractionation, Na-poor objects or their precursors appear to have reequilibrated at lower temperatures, which raised the bulk Na content of the objects, but not to the levels seen in Na-rich objects. Enrichments of Fe2+, Na, K, and P on the margins of some Na-poor objects suggest that they partly reacted with volatile-rich surroundings, both before and after brecciation. Sodium-rich objects are characterized by (1) plagioclase or glass with oligoclase or albite composition (An2-25), (2) roughly chondritic Na/Al ratios and volatile-element (Na, K, Mn) abundances, and (3) bulk-chemical trends similar to those shown by melt-pocket glasses in ordinary chondrites. Sodium-rich melt objects could have formed by the shock-melting of chondritic precursors. Feldspathic compositions for some Na-rich melt objects can be explained by preferential shock-melting of feldspar or feldspathic glass in chondritic target materials. Literature data imply that the same two chemical populations of objects, Na-poor and Na-rich, occur among smaller, normal-sized chondrules in Type 3 ordinary chondrites, suggesting that the same processes that affected large melt objects also affected chondrules.

Petrogenesis of silicate inclusions in the Weekeroo Station IIE iron meteorite: Differentiation, remelting, and dynamic mixing

The Weekeroo Station IIE iron meteorite contains a variety of felsic and mafic inclusions enclosed in an FeNi-metal host. Petrographic, EMP, and SIMS data suggest that the petrogenesis of the silicates was complex, and included differentiation, remelting, FeO-reduction, and dynamic mixing of phases. Differentiation produced a variety of olivine-free inclusion assemblages, ranging from pyroxene 1 plagioclase 1 tridymite with peritectic compositions, to coarse orthopyroxene, to plagioclase 1 tridymite and its glassy equivalent. Individual phases have similar trace-element abundances and patterns, despite large variations in inclusion textures, modes, and bulk compositions, probably as a result of mechanical separation of pre-existing phases in an impact event that dynamically mixed silicates with the metallic host. Trace- element data imply that augite and plagioclase grains in different inclusions crystallized from the same precursor melt, characterized by relatively unfractionated REE abundances of ;20 -30 3 CI-chondrites except for a negative Eu anomaly. Such a precursor melt could have been produced by ;2-5% equilibrium partial melting of an H-chondrite silicate protolith, or by higher degrees of partial melting involving subsequent fractional crystallization. Glass appears to have formed by the remelting of pre-existing plagio- clase and orthopyroxene, in a process that involved either disequilibrium or substantial melting of these phases. During remelting, silicate melt reacted with the FeNi-metal host, and FeO was reduced to Fe-metal. Following remelting and metal-silicate mixing, inclusions apparently cooled at different rates in a near-surface setting on the parent body; glass- or pigeonite-bearing inclusions cooled more rapidly (

Hf–W, Sm–Nd, and Rb–Sr isotopic evidence of late impact fractionation and mixing of silicates on iron meteorite parent bodies

2.5°C/hr between 1000 - 850°C) than pigeonite-free, largely crystalline inclusions. The results of this study point to two likely models for forming IIE iron meteorites, both involving collision between an FeNi-metal impactor and either a differentiated or undifferentiated silicate-rich target of H- chondrite affinity. Each model has difficulties and it is possible that both are required to explain the diverse IIE group. Copyright

Olivine coronas, metamorphism, and the thermal history of the Morristown and Emery mesosiderites

Abstract We report the first Sm–Nd and Rb–Sr isotopic analyses of silicate inclusions in four IIE iron meteorites: Miles, Weekeroo Station A and B, and Watson. We also report the Hf–W isotopic composition of a silicate inclusion from Watson and 182 W/ 184 W of the host FeNi metal in all four IIEs. The host metal in Watson has a negative ϵ W value (−2.21±0.24), similar to or higher than other iron meteorites [1,35] and consistent with segregation of metal from silicate early in solar system history. However, the large silicate inclusion in the Watson IIE iron yielded a chondritic ϵ W value (−0.50±0.55), thus indicating a lack of equilibration with the FeNi host within the practical lifetime of activity of the parent 182 Hf (∼50 Ma). One of the silicate inclusions in Miles is roughly chondritic in major-element composition, has a present-day ϵ Nd of +10.3, relatively non-radiogenic 87 Sr/ 86 Sr (0.714177±13), and a T CHUR age of 4270 Ma. Two silicate inclusions from Weekeroo Station and one from Watson exhibit fractionated Sm/Nd and Rb/Sr ratios, and more radiogenic 87 Sr/ 86 Sr (0.731639±12 to 0.791852±11) and non-radiogenic ϵ Nd values (−5.9 to −13.4). The silicate inclusion in Watson has a T CHUR age of 3040 Ma, in agreement with previously determined 4 He and 40 Ar gas retention ages, indicative of a late thermal event. A later event is implied for the two silicate inclusions in Weekeroo Station, which yield indistinguishable T CHUR ages of 698 and 705 Ma. Silicate inclusions in IIE iron meteorites formed over a period of 3 billion yr by impacts, involving an H-chondrite parent body and an FeNi metal parent body. The LILE-enriched nature of some of these silicates suggests several stages of melting, mixing, and processing. However, there is little evidence to suggest that the silicates in the IIE irons were ever in equilibrium with the host FeNi metal.

Mineral layers around coarse‐grained, Ca‐Al‐rich inclusions in CV3 carbonaceous chondrites: Formation by high‐temperature metasomatism

Abstract Coronas are present on all millimeter-sized mineral clasts of olivine in the Emery and Morristown mesosiderites and are a manifestation of high-temperature (T ≈ 850–1100°C) metamorphism. These coronas formed by reaction and diffusion between olivine and a mesosiderite-like matrix assemblage. The bulk composition of the coronas can be approximated by a mixture of ≈ 10–25 wt% olivine and as 90–75 wt% metal-free matrix, except for P and Cr, which are significantly enriched in coronas. Phosphorus and Cr diffused relatively rapidly to coronas and were derived from a large volume of matrix, most likely from metal that was originally enriched in these elements prior to metamorphism. The coronas in both meteorites show a similar zone sequence, but are systematically thicker in Emery (≈800 μm wide) than in Morristown (≈350 μm wide), suggesting that Emery experienced more grain growth and more intensive metamorphism than Morristown. Textural relationships suggest that corona formation and high-temperature metamorphism occurred largely after intensive millimeter-scale brecciation and after or during metal-silicate mixing. A local equilibrium model can explain many features of the coronas, but chemical equilibrium was maintained only on a very small scale. Overgrowths are present on plagioclase in the coronas of both mesosiderites and probably formed during high-temperature metamorphism. The compositional interface between core and overgrowth plagioclase is extremely sharp, suggesting that cooling rates were ≥0.1°C/y at the peak temperature of metamorphism, consistent with high-temperature metamorphism occurring in a near-surface region of the parent body.

GIANT IMPACT AND FISSION HYPOTHESES FOR THE ORIGIN OF THE MOON : A CRITICAL REVIEW OF SOME GEOCHEMICAL EVIDENCE

Coarse-grained (Type A, B) Ca-Al-rich inclusions (CAIs) in carbonaceous chondrites typically are surrounded by thin mineral layers (“rims”) that have puzzled researchers for two decades. Quantitative reaction-diffusion models can account for the overall mineral zoning structures of rims and the major-element zoning of the ubiquitous clinopyroxene layer, suggesting that the layers formed by metasomatism. Melilite-bearing CAIs appear to have reacted with an external medium that primarily contained Mg-Si-rich vapor (with atomic Mg/[Mg+Si] ≤0.66) and forsteritic olivine. Different reactant compositions in the external medium appear to have been largely responsible for producing different rim types. Various rims formed either in different local environments or at different times in an evolving system. It is suggested that layer formation occurred in a nebular setting, while silicates were being vaporized and olivine was condensing around CAIs. Steady state layer growth models do not adequately explain the presence of melilite layers or patches in some rims and consistently underestimate the spinel/clinopyroxene ratios of rims, probably because of a failure to attain complete steady state conditions as a result of changing pressure, temperature, or reactant compositions during layer growth. Roughly 3–50% of the spinel in rims can be attributed to metasomatic growth, but the remaining spinel formed by another process, possibly as a residue of partial melting during a brief vaporization event, or by preferential nucleation on the surfaces of molten CAIs. The thermal events accompanying CAI metasomatism can be constrained by modeling Mg isotope exchange that occurred between some CAIs and the external medium. Based on one well-studied CAI, it is inferred that isotopic exchange and layer formation was initiated either in a high-temperature (>1450°C) heating event <10 hours in duration, or at lower temperatures (≤1450°C) during cooling at a rate of ≤0.1–2°C/hr.

Shock-induced mobilization of metal and sulfide in planetesimals: Evidence from the Buck Mountains 005 (L6 S4) dike-bearing chondrite

Geochemical data for Ni, Co, Cr, V, and Mn have played an important role in theories for the Moons origin. It has been argued that the data for these elements strongly support formation of the Moon as ejecta from the Earth, either as a result of one giant or numerous smaller impacts on the proto-Earth. These theories have come to be known as the “Giant Impact” and “Impact-triggered Fission” hypotheses, respectively, and the first of these has been the leading explanation for the origin of the Moon over the past decade. Data for these same “diagnostic” elements also have been used to argue for significant distinctions between the bulk compositions of the Moon and a eucrite (HED) parent body, which otherwise appear to be remarkably similar in their compositions. We review geochemical evidence pertaining to the origin of the Moon, focusing on the diagnostic elements, and find that there is no strong geochemical support for either the Giant Impact or Impact-triggered Fission hypotheses. We show that basalts ...

Response to the comment by G. Dreibus and H. Wanke on "Comparative geochemistry of basalts from the Moon, Earth, HED asteroid, and Mars: Implications for the origin of the Moon" (2001) Geochim. Cosmochim. Acta 65, 979 -997

Abstract The conditions under which metal cores formed in silicate-metal planetary bodies in the early Solar System are poorly known. We studied the Buck Mountains 005 (L6) chondrite with serial sectioning, X‑ray computed microtomography, and optical and electron microscopy to better understand how metal and troilite were redistributed as a result of a moderately strong (shock stage S4) shock event, as an example of how collisional processes could have contributed to differentiation. The chondrite was recovered on Earth in multiple small pieces, some of which have a prominent, 1.5-3 mm wide holocrystalline shock melt dike that forms a jointed, sheet-like structure, as well as an associated shock vein network. The data suggest that metal and troilite within the dike were melted, sheared, and transported as small parcels of melt, with metal moving out of the dike and along branching veins to become deposited as coarser nodules and veins within largely unmelted host. Troilite also mobilized but partly separated from metal to become embedded as finer-grained particles, vein networks, and emulsions intimately intergrown with silicates. Rock textures and metal compositions imply that shock melts cooled rapidly against relatively cool parent body materials, but that low-temperature annealing occurred by deep burial within the parent body. Our results demonstrate the ability of shock processes to create larger metal accumulations in substantially unmelted meteorite parent bodies, and they have implications for the formation of iron meteorites and for core formation within colliding planetesimals.

Phase Analysis of Large EDS Datasets with Matlab

The central theses of our work (Ruzicka et al 1998, 2001) are that the composition of the Moon is not unique in the solar system, that it resembles the composition of the parent body of HED meteorites (likely the asteroid 4-Vesta) to a remarkable extent, and that geochemical data do not support an origin of the Moon by rotational fission (Binder, 1986) or small-impact collisional ejection (Ringwood, 1986, 1990). Furthermore, the data are consistent with a giant-impact origin for the Moon (Hartmann and Davis, 1975; Hartmann, 1986; Cameron, 1986, 1997, 2000; Canup and Asphaug, 2001) only if the Moon largely inherited the composition of the impactor. Few researchers dispute the idea that the Moon and HED parent body have similar abundances of many volatile (e.g., Na and K) and siderophile elements (e.g., W). The controversy (Dreibus and Wanke, 2002) centers instead on the interpretation of geochemical data for a few key elements, including Ni, Co, Cr, and Mn. According to Wanke and colleagues, for these elements the composition of the Moon and Earth are similar, but very different for the Moon and HED parent body (Dreibus et al 1977, 1979; Wanke and Dreibus, 1986). Various authors have reached similar conclusions for Cr and Mn (Drake et al 1989; Ringwood, 1989; Ringwood et al 1990; O’Neill 1991) and for Ni and Co (Ringwood and Seifert, 1986; Ringwood, 1989; O’Neill, 1991), using either the modelled abundances of Wanke and colleagues, or approaches similar to these researchers. In contrast, based on correlated (albeit non-linear) abundances for these elements, we argue that Ni, Co and Cr abundances are similar for the Moon and HED asteroid, and different for the Moon and Earth. In contrast to Wanke and Dreibus (1986), we also argue that Mn data do not necessarily imply special conditions for the formation of the Earth and Moon. The different interpretations stem in part from the different methodologies and samples used. We based conclusions on geochemical data for analogous basaltic samples from each planetary body, and inferred the compositions of the source regions, which could correspond to large portions of the mantles of the parent objects. For lunar samples we concentrated on mare basalts, as opposed to KREEP basalts or aluminous basalts. This was because the latter appear to represent unusual differentiates and/or impact-melt rocks, which are not analogous to the basalts from other objects. In contrast, Wanke and colleagues based their conclusions on derivative compositions of the Moon, HED asteroid, Earth, and Mars, estimated from models that differed from one object to the other, that varied depending on the element being considered, and that were based on different types of rocks, including impact breccias, volcanic rocks, and plutonic rocks. The Ni, Co, Cr, and Mn abundances modelled for the HED asteroid (Dreibus et al 1977; Dreibus and Wanke, 1980) depend strongly on the assumed composition and proportion of an olivine-rich mantle component that has not yet been sampled. This olivine-rich component could have a composition similar to lunar dunite (Dreibus et al 1977), pallasitic olivine (Dreibus and Wanke, 1980), or something else altogether. The proportion of this olivine-rich component is also uncertain as it depends on the nature of the protolith. Dreibus et al (1977) and Dreibus and Wanke (1980) assumed a protolith similar to CI-chondrites, but significantly different parent body compositions are obtained for protoliths similar to ordinary chondrites or enstatite chondrites (Ruzicka et al 1997). A different choice for the composition and proportion of the HED mantle component will change the model Ni, Co, Cr, and Mn abundances in the HED asteroid. For the Moon, the bulk composition inferred by Wanke and coworkers was inferred mainly from the composition of highland breccias, which are acknowledged to be contaminated by meteoritic debris. The abundances of these elements in the Moon were derived by making assumptions about the proportion and composition of the components (KREEP, anorthosite, Mg-rich component) out of which the breccias are assumed composed, and depend on the composition of the assumed protolith (e.g., CI-chondrites) (Wanke et al 1977a, 1977b, 1978). Incorrect assumptions about any of these will lead to an incorrect derivative composition for the Moon. For Ni and Co, the bulk composition of the Moon estimated by Wanke et al (1978) was obtained only after additional assumptions were made regarding the amount and type of meteoritic contamination. Whether or not these models lead to a valid bulk lunar composition is debatable. In contrast, we argue that the compositions of eucrite and mare basalts themselves show good evidence for being controlled by igneous fractionation of mafic minerals (mainly olivine and pyroxene) from protoliths with similar Ni, Co, and Cr abundances. Although the Ni and Co abundances of eucrites are lower than that in mare basalts, the Ni-MgO and Co(MgO FeO) systematics for eucrites and all mare basalts fall on the same trend, and this trend is displaced from the compositions shown by terrestrial basalts and komatiites. On plots * Author to whom correspondence should be addressed (ruzickaa@pdx.edu). Pergamon Geochimica et Cosmochimica Acta, Vol. 66, No. 14, pp. 2633–2635, 2002 Copyright