Search

Accepted: 2017-05-26, 資料番号: SA1170244000

One way to better understand processes related to chondrite metamorphism is to evaluate changes in chondrite features as a function of petrologic type. Toward this end the valence and coordination of Ti in olivine and pyroxene in suites of ordinary (H, L, and LL) and enstatite (EH and EL) chondrites of types 3 through 6 have been determined with XANES spectroscopy. Trivalent Ti, typically 10–40% of the Ti in the analytical volumes, was found in ordinary chondrites of all types, despite the stability of oxidized iron in the samples. Average valences and the proportions of Ti that are in tetrahedral coordination generally decrease with increasing grade between types 3.0 and 3.5, increase from 3.5 to 4, and then level off. These trends are consistent with previous studies of chondrite oxidation states using other methods, except here the onset of oxidation is observed at a lower type, ∼3.5, than previously indicated (4). These results are also consistent with previous suggestions that oxidation of higher-grade ordinary chondrite samples involved exposure to aqueous fluids from melting of accreted ice. In the enstatite chondrites, typically 20–90% of the Ti is trivalent Ti, so it is reduced compared to Ti in the ordinary chondrites. Valence decreases slightly from petrologic type 3 to 4 and increases from 4 to 6, but no increases in tetrahedral coordination with petrologic type are observed, indicating a redox environment or process distinct from that of ordinary chondrite metamorphism. The presence of Ti4+ in the E chondrites supports previous suggestions that they formed from oxidized precursors that underwent reduction. Unlike ordinary chondrites, enstatite chondrites are thought to have been derived from a body or bodies that did not accrete ice, which could account for their different valence-coordination-petrologic type relationships. The hypothesis, based on observations of unmetamorphosed chondrules and supported by laboratory experiments, that equilibration of Ti valence is sluggish compared to that of Fe could account for the coexistence of reduced Ti and oxidized Fe seen in chondrites of all petrologic types.

A coordinated, electron-backscatter-diffraction (EBSD) and transmission electron microscope (TEM) study was undertaken to obtain information on the origin of rims on refractory inclusions in the Allende and Axtell CV3 chondrites. These measurements were supported by theoretical modeling using density functional theory. Crystal-orientation analysis of Wark-Lovering rims via EBSD revealed pyroxene grains with similar crystallographic orientations to one another in both inclusions. An epitaxial relationship between grains within the diopside and anorthite rim layers was observed in Allende. TEM examination of the rims of both samples also revealed oriented crystals at depth. The microstructural data on the rims suggest that grain clusters grew in the form of three-dimensional islands. Density functional theory calculations confirm that formation of oriented grain islands is the result of energy minimization at high temperature. The results point toward condensation as the mode of origin for the rims studied here.

The silicate and metal clasts in CB chondrites have been inferred to form as condensates from an impact-generated vapor plume between a metal-rich body and a silicate body. A detailed study of the condensation of impact-generated vapor plumes showed that the range of CB silicate clast compositions could not be successfully explained without invoking a chemically differentiated target. Here, we report the most comprehensive elemental study yet performed on CB silicates with 32 silicate clasts from nine slices of Gujba analyzed by laser ablation inductively coupled plasma mass spectrometry for 53 elements. Like in other studies of CBs, the silicate clasts are either barred olivine (BO) or cryptocrystalline (CC) in texture. In major elements, the Gujba silicate clasts ranged from chondritic to refractory enriched. Refractory element abundances ranged from 2t o 10� CI, with notable anomalies in Ba, Ce, Eu, and U abundances. The two most refractory-enriched BO clasts exhibited negative Ce anomalies and were depleted in U relative to Th, characteristic of volatilization residues, while other BO clasts and the CC clasts exhibited positive Ce anomalies with excess U (1–3 � CI), and Ba (1–6 � CI) anomalies indicating recondensation of ultra-refractory element depleted vapor. The Rare Earth Elements (REE) also exhibit light REE (LREE) enrichment or depletion in several clasts with a range of (La/Sm)CI of 0.9–1.8. This variation in the LREE is essentially impossible to accomplish by processes involving vapor–liquid or vapor–solid exchange of REE, and appears to have been inherited from a differentiated target. The most distinctive evidence for inherited chemical differentiation is observed in highly refractory element (Sc, Zr, Nb, Hf, Ta, Th) systematics. The Gujba clasts exhibit fractionations in Nb/Ta that correlate positively with Zr/Hf and span the range known from lunar and Martian basalts, and exceed the range in Zr/Hf variation known from eucrites. Variations of highly incompatible refractory elements (e.g., Th) against less incompatible elements (e.g., Zr, Sr, Sc) are not chondritic, but exhibit distinctly higher Th abundances requiring a differentiated crust to be admixed with depleted mantle in ratios that are biased to higher crust/mantle ratios than in a chondritic body. The possibility that these variations are due to admixture of refractory inclusion-debris into normal chondritic matter is raised but cannot be definitively tested because existing ‘‘bulk” analyses of CAIs carry artifacts of unrepresentative sampling. The inferences drawn from the compositions of Gujba silicate clasts, here, complement what has been inferred from the compositions of metallic clasts, but provide surprisingly detailed insight into the structure of the target. Evidence that metal and silicate in CB chondrites both formed from impact-generated vapor plumes, taken together with recent work on metallic nodules in E chondrites, and on ordinary chondrites, indicates that chondrule formation occurs by this mechanism quite widely. However, the nature of

The impact hypothesis for the origin of CB chondrites was tested by performing equilibrium condensation calculations in systems composed of vaporized mixtures of projectile and target materials. When one of the impacting bodies is composed of the metal from CR chondrites and the other is an H chondrite, good agreement can be found between calculated and observed compositions of unzoned metal grains in CB chondrites but the path of composition variation of the silicate condensate computed for the same conditions that reproduce the metal grain compositions does not pass through the measured compositions of barred olivine (BO) or cryptocrystalline (CC) chondrules in the CBs. The discrepancy between measured chondrule compositions and those of calculated silicates is not reduced when diogenite, eucrite or howardite compositions are substituted for H chondrite as the silicate-rich impacting body. If, however, a CR chondrite body is differentiated into core, a relatively CaO-, Al 2 O 3 -poor mantle and a CaO-, Al 2 O 3 -rich crust, and later accretes significant amounts of water, a collision between it and an identical body can produce the necessary chemical conditions for condensation of CB chondrules. If the resulting impact plume is spatially heterogeneous in its proportions of crust and mantle components, the composition paths calculated for silicate condensates at the same P tot , Ni/H and Si/H ratios and water abundance that produce good matches to the unzoned metal grain compositions pass through the fields of BO and CC chondrules, especially if high-temperature condensates are fractionated in the case of the CCs. While equilibrium evaporation of an alloy containing solar proportions of siderophiles into a dense impact plume is an equally plausible hypothesis for explaining the compositions of the unzoned metal grains, equilibrium evaporation can explain CB chondrule compositions only if an implausibly large number of starting compositions is postulated. Kinetic models applied to co-condensing metal grains and silicate droplets in a region of the plume with very similar composition, but with high cooling rate and sharply declining P tot during condensation, produce very good matches to the zoning profiles of Ir, Ni, Co and Cr concentrations and Fe and Ni isotopic compositions observed in the zoned metal grains in CB chondrites but produce very large positive δ 56 Fe in the cogenetic silicate, which are not found in the chondrules.

Ti valence measurements in MgAl_2O_4 spinel from calcium-aluminum-rich inclusions (CAIs) by X-ray absorption near-edge structure (XANES) spectroscopy show that many spinels have predominantly tetravalent Ti, regardless of host phases. The average spinel in Allende type B1 inclusion TS34 has 87% Ti^(+4). Most spinels in fluffy type A (FTA) inclusions also have high Ti valence. In contrast, the rims of some spinels in TS34 and spinel grain cores in two Vigarano type B inclusions have larger amounts of trivalent titanium. Spinels from TS34 have approximately equal amounts of divalent and trivalent vanadium. Based on experiments conducted on CAI-like compositions over a range of redox conditions, both clinopyroxene and spinel should be Ti^(+3)-rich if they equilibrated with CAI liquids under near-solar oxygen fugacities. In igneous inclusions, the seeming paradox of high-valence spinels coexisting with low-valence clinopyroxene can be explained either by transient oxidizing conditions accompanying low-pressure evaporation or by equilibration of spinel with relict Ti^(+4)-rich phases (e.g., perovskite) prior to or during melting. Ion probe analyses of large spinel grains in TS34 show that they are enriched in heavy Mg, with an average Δ^(25)Mg of 4.25 ± 0.028‰, consistent with formation of the spinel from an evaporating liquid. Δ^(25)Mg shows small, but significant, variation, both within individual spinels and between spinel and adjacent melilite hosts. The Δ^(25)Mg data are most simply explained by the low-pressure evaporation model, but this model has difficulty explaining the high Ti^(+4) concentrations in spinel.

Sodium saturation of the vapor coexisting with chondrules at their liquidus temperatures implies that vapor-condensed phase equilibrium was reached at those temperatures for all elements more refractory than sodium. In order to investigate the possibility that chondrules formed in impact-generated plumes, equilibrium calculations were applied to droplets made from two different target compositions. Combinations of dust enrichment and P tot were found that lead to sodium saturation, and the subsequent chemical and mineralogical evolution of the droplets was explored at those conditions. If an impact on a body of CI composition caused instantaneous heating, melting and devolatilization of the target rock and ejection of a plume of gaseous, liquid and solid matter that mixed with residual nebular gas at conditions where 50% or 90% of the sodium was retained by the resulting droplets at their liquidus temperature, their mineralogical and chemical properties would strongly resemble those of Type II chondrules. If the droplets cooled and equilibrated with the mixture of residual nebular gas and their devolatilized water, sulfur and alkalis, the fayalite content of the olivine and the chemical compositions of the bulk droplets and their glasses would closely resemble those of Types IIA and IIAB chondrules at CI dust enrichments between 400× and 800×. For 50% sodium retention, the corresponding values of P tot are 2 bars (for 400×) and 1 bar (for 800×). For 90% retention, they are 25 and 10 bars, respectively. If, instead, the target has an anhydrous, ordinary chondrite-like composition, called H′, the ejected droplets are bathed in a gas mix consisting mostly of devolatilized sulfur and alkalis with residual nebular gas, a much more reducing plume. If the conditions were such that sodium were retained by the resulting droplets at their liquidus temperature, the fayalite contents of the olivine and the chemical compositions of the bulk droplets and their glasses would closely resemble those of Types IA and IAB chondrules at H′ dust enrichments between 10 3 × and 4 × 10 3 ×. For 90% sodium retention, the corresponding values of P tot are 15 bars (for 10 3 ×) and 2 bars (for 4 × 10 3 ×). For 50% retention, they are 2 and 8 × 10 −2 bars, respectively.

We present high-precision measurements of the Mg isotopic compositions of a suite of types I and II chondrules separated from the Murchison and Murray CM2 carbonaceous chondrites. These chondrules are olivine- and pyroxene-rich and have low 27 Al/ 24 Mg ratios (0.012-0.316). The Mg isotopic compositions of Murray chondrules are on average lighter (d 26 Mg ranging from 0.95& to 0.15& relative to the DSM-3 standard) than those of Murchison (d 26 Mg ranging from 1.27& to +0.77&). Taken together, the CM2 chondrules exhibit a narrower range of Mg isotopic compositions than those from CV and CB chondrites studied previously. The least-altered CM2 chondrules are on average lighter (average d 26 Mg = 0.39 0.30&, 2SE) than the moderately to heavily altered CM2 chondrules (average d 26 Mg = 0.11 0.21&, 2SE). The compositions of CM2 chondrules are consistent with isotopic fractionation toward heavy Mg being associated with the formation of secondary silicate phases on the CM2 parent body, but were also probably affected by volatilization and recondensation processes involved in their original formation. The low-Al CM2 chondrules analyzed here do not exhibit any mass-independent variations in 26 Mg from the decay of 26 Al, with the exception of two chondrules that show only small variations just outside of the analytical error. In the case of the chondrule with the highest Al/Mg ratio (a type IAB chondrule from Murchison), the lack of resolvable 26 Mg excess suggests that it either formed >1 Ma after calcium-aluminum-rich inclusions, or that its Al-Mg isotope systematics were reset by secondary alteration processes on the CM2 chondrite parent body after the decay of 26 Al.

– For fayalite formation times of several thousand years, and systems enriched in water by a factor of ten relative to solar composition, 1 μm radius olivine grains could reach 2 mole% fayalite and 0.1 μm grains 5 mole% by nebular condensation, well short of the values appropriate for precursors of most chondrules and the values found in the matrices of unequilibrated ordinary chondrites. Even 10 μm olivine crystals could reach 30 mole% fayalite above 1100 K in solar gas if condensation of metallic nickel-iron were delayed sufficiently by supersaturation. Consideration of the surface tensions of several phases with equilibrium condensation temperatures above that of metallic iron shows that, even if they were supersaturated, they would still nucleate homogeneously above the equilibrium condensation temperature of metallic iron. This phenomenon would have provided nuclei for heterogeneous nucleation of metallic nickel-iron, thus preventing the latter from supersaturating significantly and preventing olivine from becoming fayalitic. Unless a way is found to make nebular regions far more oxidizing than in existing models, it is unlikely that chondrule precursors or the matrix olivine grains of unequilibrated ordinary chondrites obtained their fayalite contents by condensation processes. Perhaps stabilization of FeO occurred after condensation of water ice and accretion of icy planetesimals, during heating of the planetesimals and/or in hot, dense, water-rich vapor plumes generated by impacts on them. This would imply that FeO is a relatively young feature of nebular materials.

When a shock wave passes through a nebular gas, increasing water enrichment leads to higher temperatures and postshock P tot , but lower cooling rates. A kinetic evaporation model is developed for tracking the chemical and isotopic changes that would occur in a clump of chondrule precursor dust surrounded by nebular gas in a closed system traversed by a nebular shock wave, taking into account effects of non-equilibrium melting and fractional crystallization on the liquid composition and the temperature difference between the gas and the droplet. A range of shock wave temperature–pressure histories computed for systems enriched relative to solar composition by factors of 550 in water, to achieve the redox state of chondrules, and 600 in dust, to retard evaporation, are employed, and redox changes are assumed to occur on the time–scale of heating and cooling in each. Two different system compositions are assumed, with the mean Fe/Si ratios of Types I and II chondrules. Two different textural outcomes are modeled, PO, in which nuclei are preserved and olivine crystallization begins immediately upon reaching saturation, and BO, in which no nuclei are preserved and olivine crystallization begins only after 300–400 K of supersaturation. In all cases, all iron evaporates, regardless of its oxidation state, as well as alkalis and smaller fractions of Mg and Si. In most cases, recondensation occurs on the time-scale of cooling, resulting in droplets whose bulk compositions have small isotopic anomalies in Mg, Si and Fe, comparable to those seen in bulk chondrules. Because fractional crystallization of olivine occurs before recondensation is complete, however, large isotopic variations, especially for iron, would have been recorded both within olivine crystals and between olivine and glass within these objects. Even after diffusive relaxation during crystal growth and cooling, variations in d 25 Mg of several tenths of a & to several & ,i nd 29 Si of 0.1& to several & and in d 56 Fe of several & would be measurable within large grains that grew throughout the olivine crystallization interval in many cases, and olivine-glass differences of Pseveral tenths of a & in d 29 Si, and of several & in d 56 Fe would be preserved. Such internal isotopic heterogeneities have not yet been observed in chondrules, suggesting that the latter did not form in these shock wave thermal histories. Suppression of production of internal isotopic variations requires heating times that are shorter by a factor of 100, combined with dust enrichments P 6 � 10 4 and/or P tot P10 � 2 bar. Together with relatively high fO2 , these

– Acfer 094 is an unshocked, nearly unaltered carbonaceous chondrite with an unusual suite of refractory inclusions. The refractory inclusions in a newly prepared thin section and a small aliquot of disaggregated material were studied to compare the population with previous work, and to report new or unusual inclusion types. A total of 289 Ca-, Al-rich inclusions in the thin section and 67 among the disaggregated material, having a total of 31 different mineral assemblages, were found. Inclusions are largely free of secondary alteration products, and are typically ≤200 μm across. The most common are gehlenitic melilite+spinel±perovskite, spinel+perovskite, and spinel with a thin, silicate rim, typically melilite±diopside. Such rims and (thicker) mantles are very common among Acfer 094 inclusions, and they exhibit a variety of zoning patterns with respect to akermanite and FeO contents. In the thin section, about 13% of the inclusions contain hibonite and approximately 5% are grossite-bearing; in the disaggregated material, the percentages are 14 and 9, respectively, comparable to previous work. Among the unusual inclusions are a fine-grained, porous, Ti-rich hibonite+spinel+perovskite+melilite inclusion with a compact, coarse, Ti-poor hibonite+spinel+melilite clast; two inclusions in which hibonite has reacted to form grossite; two inclusions with FeO-rich spinel; and a small object consisting of fassaite enclosing euhedral spinel, the first fragment of a Type B inclusion reported from Acfer 094. Inclusions similar to those found in CM or CV chondrites are rare; Acfer 094 contains a distinctive population of inclusions. The population, dominated by small, melilite-bearing inclusions, is most similar to that of CO chondrites. A distinguishing feature is that in Acfer 094, almost every phase in almost every refractory inclusion contains 0.5–1.5 wt% FeO. A lack of diffusion gradients and the pristinity of the matrix imply that the inclusions experienced prolonged exposure to FeO-bearing fluid prior to accretion into the Acfer 094 parent body. There are no known nebular conditions under which the refractory phases found in the present samples could acquire FeO enrichments to the observed levels. The most likely setting is therefore in an earlier, FeO-rich parent body. The inclusions were ejected from this parent body, mixed with typical CAIs, chondrules, amoeboid olivine aggregates, and amorphous material, and incorporated into the Acfer 094 parent body.

Calcium Tschermak’s pyroxene (CaTs), CaAlAlSiO_6, is well known as an important component in pyroxene. It is a member of the Ca clinopyroxene group in which Al dominates in the M1 site. Pyroxenes with more than 80 mol% CaTs were observed previously in Ca-,Al-rich refractory inclusions (CAI) from five carbonaceous chondrites. This study re-investigated the near end-member CaTs in the Allende and Murray chondrites. Electron backscatter diffraction (EBSD) is used to establish that its crystal structure is monoclinic, C2/c; a = 9.609 A, b = 8.652 A, c = 5.274 A, β =106.06°, V = 421.35 A^3, and Z = 4. Its EBSD pattern is an excellent match to that of synthetic CaAlAlSiO_6 with the C2/c structure. MicroRaman is also carried out to confirm the crystal structure. The Allende CaTs, with 46.00 wt% Al_(2)O_3 and 97 mol% Al in the M1 site, has the formula Ca_(1.02)(Al_(0.97)Fe_(0.01)Mg_(0.01))0.99(Si_(1.00)Al_(1.00))_(Σ2.00)O_6. It occurs as micrometer-sized crystals along with melilite, hibonite, perovskite, spinel, corundum, Ti^(3+)-rich pyroxene, and grossular in a fluffy Type A CAI. It is probably a secondary phase resulting from the alteration of gehlenitic melilite. The CaTs in Murray, with a formula Ca_(0.98)(Al_(0.81)Mg_(0.16)Ti_(0.04)^(4+))_(Σ1.01)(Si_(1.11)Al_(0.89))_(Σ2.00)O_6, occurs with hibonite and Al-rich diopside in a glass-free refractory spherule. This sample formed by solidification of a once-molten droplet early in the history of the solar system.

We compare the observed composition ranges of olivine, pyroxene, and Fe-Ni sulfides in Wild 2 grains with those from chondritic interplanetary dust particles (IDPs) and chondrite classes to explore whether these data suggest affinities to known hydrous materials in particular. Wild 2 olivine has an extremely wide composition range, from Fa096, with a pronounced frequency peak at Fa1. The composition range displayed by the low-calcium pyroxene is also very extensive, from Fs48 to Fs0, with a significant frequency peak centered at Fs5. These ranges are as broad or broader than those reported for any other extraterrestrial material. Wild 2 Fe-Ni sulfides mainly have compositions close to that of FeS, with less than 2 atom% Ni; to date, only two pentlandite grains have been found among the Wild grains, suggesting that this mineral is not abundant. The complete lack of compositions between FeS and pentlandite (with intermediate solid solution compositions) suggests (but does not require) that FeS and pentlandite condensed as crystalline species, i.e., did not form as amorphous phases, which later became annealed. While we have not yet observed any direct evidence of water-bearing minerals, the presence of Ni-bearing sulfides, and magnesium-dominated olivine and low-Ca pyroxene does not rule out their presence at low abundance. We do conclude that new investigations of major- and minorelement compositions of chondrite matrix and IDPs are required.

Crystallization experiments on liquids with compositions similar to those of compact Type A, Type B1 and Type B2 refractory inclusions were conducted under controlled temperature and fO2 conditions. Application of the results to the compositions of coexisting Ti 3+ -bearing fassaitic clinopyroxene + melilite pairs in natural inclusions shows that, if they crystallized at ~1509 K, they did so at log fO2 = −19.8 ± 0.9, only slightly below the equilibrium log fO2 of a partially condensed system of solar composition at the same temperature, − −

The clinopyroxene in coarse-grained refractory inclusions contains significant amounts of Ti and V, two elements that are multivalent over the range of temperatures and oxygen fugacities under which the inclusions formed. The Ti3+/Ti4+ ratios and the valence of V of these pyroxenes are valuable recorders of nebular conditions. The former can be calculated stoichiometrically from electron probe analyses, but only for relatively Ti-rich grains (i.e., >∼4 wt% TiO 2 tot ). For Ti-poor pyroxene, and for measurement of V valence, another technique is needed. We have, for the first time, applied K-edge X-ray absorption near edge structure (XANES) spectroscopy to the measurement of Ti and V valence in meteoritic clinopyroxene in refractory inclusions. Use of MicroXANES, a microbeam technique with high (μm-scale) spatial resolution, allowed measurement of Ti and V valence along traverses across (a) Ti-, V-rich “spikes” in pyroxene in Type B1 inclusions; (b) typical grains in a Type B2 inclusion; and (c) the pyroxene layer of the Wark–Lovering rim sequence on the outsides of two inclusions. Measurements of Ti3+/(Ti3+ + Ti4+), or Ti3+/Titot, by XANES agree with values calculated from electron probe analyses to within ∼0.1, or ∼2σ. The results show that Ti3+/Titot increases sharply at the spikes, from 0.46 ± 0.03 to 0.86 ± 0.06, but the V valence, or V2+/(V2+ + V3+), does not change, with V2+ ≈ V3+. We found that pyroxene in both Types B1 and B2 inclusions has Ti3+/Titot and V2+/Vtot ratios between 0.4 and 0.7, except for the spikes. These values indicate, to first order, formation at similar, highly reducing oxygen fugacities that are consistent with a solar gas. The pyroxene in the rim on an Allende fluffy Type A coarse-grained refractory inclusion, TS24, has an average Ti3+/Titot of 0.51 ± 0.08 and an average V2+/Vtot of 0.61 ± 0.06, determined by XANES. These values are within the range of those of pyroxene in the interiors of inclusions, indicating that the rims also formed under highly reducing conditions. Measurements of Ti3+/Titot of pyroxene in the rim of a Leoville compact Type A inclusion, 144A, by both XANES and electron probe give a wide range of results. Of our 72 XANES analyses of this rim, 66% have Ti3+/Titot of 0.40–0.71, and the remaining analyses range from 0 to 0.38. In data from Simon et al. [Simon J. I., Young E. D., Russell S. S., Tonui E. K., Dyl K. A., and Manning C. E. (2005) A short timescale for changing oxygen fugacity in the solar nebula revealed by high-resolution 26Al–26Mg dating of CAI rims. Earth Planet. Sci. Lett. 238, 272–283.] for this sample, 7 electron probe analyses yield calculated Ti3+/Titot values that are positive and 15 do not. In the probe analyses that have no calculated Ti3+, Ca contents are anticorrelated and Al contents directly correlated with the total cations per 6 oxygens, and the data fall along trends calculated for addition of 1–7% spinel to pyroxene. It appears likely that electron probe analyses of pure pyroxene spots have Ti3+/Titot values that are typical of refractory inclusions, in agreement with the majority of the XANES results. The average of the XANES data for 144A, 0.41 ± 0.14, is within error of that for TS24. The rim of 144A probably formed under reducing conditions like those expected for a solar gas, and was later heterogeneously altered, resulting in an uneven distribution of secondary, FeO-, Ti-bearing alteration products in the rim, and accounting for the measurements with low Ti3+/Titot values.

We report petrography, mineral chemistry, bulk chemistry, and bulk isotopic compositions of a suite of 40 spinel-rich inclusions from the Murchison (CM2) carbonaceous chondrite. Seven types of inclusions have been identified based on mineral assemblage: spinel-hibonite-perovskite; spinel-perovskite-pyroxene; spinel-perovskite-melilite; spinel-hibonite-perovskite-melilite; spinel-hibonite; spinel-pyroxene; and spinel-melilite-anorthite. Hibonite-bearing inclusions have Ti-poor spinel compared to the hibonite-free ones, and spinel-hibonite-perovskite inclusions have the highest average bulk TiO 2 contents (7.8 wt%). The bulk CaO/Al 2 O 3 ratios of the inclusions range from 0.005 to 0.21, well below the solar value of 0.79. Hibonite-, spinel-rich inclusions consist of phases that are not predicted by condensation calculations to coexist; in the equilibrium sequence, hibonite is followed by melilite, which is followed by spinel. Therefore, hibonite-melilite or melilite-spinel inclusions should be dominant instead. One explanation for the “missing melilite” is that it condensed as expected, but was lost due to evaporation of Mg and Ca during heating and melting of spherule precursors. If this theory were correct, melilite-poor spherules would have isotopically heavy Mg and Ca, assuming Rayleigh fractionation accompanied evaporation. Except for one inclusion with F Mg = 4.3 ± 2.6‰/amu and another with isotopically light Ca ( F Ca = −3.4 ± 2.0‰/amu), however, all the inclusions we analyzed have normal isotopic compositions within their 2σ uncertainties. Thus, we found no evidence for significant mass-dependent fractionation. Conditions necessary for non-Rayleigh evaporation are unlikely if not unrealistic, and our preferred explanation for the general lack of melilite among hibonite-, spinel-bearing inclusions is kinetic inhibition of melilite condensation relative to spinel. Because of similarities between the crystal structures of hibonite and spinel, it should be easier for spinel than for melilite to form from hibonite.

Most of the petrologic data available for Type B inclusions comes from Type B1s. Relatively little comes from the B2s, and there has not been a systematic comparison of the properties of their two most abundant minerals. In this work, we document the compositions and zoning patterns of melilite and fassaite in Type B2 inclusions, and compare and contrast them with the features of their counterparts in Type B1 inclusions. We find that melilite compositions in Type B2 inclusions are similar to those of Type B1s, with maximum Ak contents of ∼75 mol % and a positive correlation between Ak and Na 2 O contents. Asymmetrically zoned melilite is common in Type B2s as are melilite grains with reversely zoned regions, and the reversely zoned portions of crystals are thicker than in B1s. In B2s, like B1s, fassaite is zoned with decreasing Ti, Sc, and V oxide contents from cores to rims of grains. Approximately half of the Ti is trivalent, but unlike that in B1s, within fassaite grains in B2s the Ti 3+ /(Ti 3+ + Ti 4+ ) ratio does not decrease from core to rim, and sharp enrichments (“spikes”) in Ti 3+ and V are not observed. Sector-zoned fassaite is much more common in B2s than in B1s. The differences we observed can be accounted for by the differences in bulk compositions between B1s and B2s. Type B2 inclusions tend to have higher SiO 2 contents, hence higher An/Ge component ratios, than Type B1s. Phase equilibria show that, compared to B1s, in B2s less melilite should crystallize prior to the appearance of fassaite, so that in B2s a higher proportion of melilite cocrystallizes with fassaite, causing more of the crystals to be reversely zoned; more melilite crystallizes while adjacent to other crystals, leading to asymmetrical zoning; and with more liquid available, transport of components to growing fassaite occurs more readily than in B1s, facilitating crystal growth and giving rise to sector zoning. The lack of zoning with respect to Ti 3+ /Ti tot and the absence of Ti 3+ -, V-rich spikes suggest that Type B2 melts maintained equilibrium with the nebular gas throughout crystallization, while the interiors of B1s were probably isolated from the gas, perhaps by their melilite mantles. This makes the similarity of Na-Ak relationships in B1 and B2 melilite difficult to understand, but apparently enclosure by melilite mantles was not necessary for the retention of Na 2 O during crystallization of Type B refractory inclusions.

A large (7 mm in diameter) Allende type B inclusion has a typical bulk composition and a unique structure: a fassaite-rich mantle enclosing a melilite-rich core. The core and mantle have sharply contrasting textures. In the mantle, coarse (~1 mm across), subhedral fassaite crystals enclose radially oriented melilite laths about 500 m long that occur at the inclusion rim. The core consists of blocky melilite grains 20-50 μm across and poikilitically enclosed in anhedral fassaite grains that are optically continuous over ~1 mm. Another unique feature of this inclusion is that melilite laths also extend from the core into the mantle. Fassaite in both the core and mantle is very rich in fine-grained (1-10 μm) spinel. The rim laths are normally zoned (Ak 30-70) inward from the rim of the inclusion with reverse zoning over the last ~200 m to crystallize. A very wide range of melilite compositions is found in the core of the inclusion, where gehlenitic grains (Ak 5-12) occur. These grains are enclosed in strongly zoned (Ak 15-70) overgrowths. The gehlenitic cores and innermost parts of the overgrowths are Na2O-free, but the outer parts of the overgrowths are not. In the laths at the rim, Na2O decreases inward from the rim, then increases. Fassaite in the core has the same range of Ti contents as that in the mantle: 29 wt% TiO2 + Ti2O3. Two melting events are required to account for the features of this inclusion. In the first event, the precursor assemblage is heated to ~1400 °C and melts except for gehlenitic (Ak 5-12) melilite and some spinel. These grains become concentrated in the core. During cooling, Na2O-free melilite nucleates at the rim of the inclusion and on the relict grains in the core. After open system secondary alteration, the inclusion is heated again, but only to ~1260 °C. Melilite more gehlenitic than k40 does not melt. During cooling, Na2O-bearing melilite crystallizes as small, blocky grains and laths in the core and as overgrowths on relict grains in the core and at the rim. Eventually melilite co-crystallizes with fassaite, leading to the reverse zoning observed in the laths. The coexistence in this inclusion of Na-free and Na-bearing melilite, plus a positive correlation between Na2O and kermanite contents in melilite in an inclusion with a bulk Mg isotopic composition that is mass-fractionated in favor of the heavy isotopes, are both consistent with at least two melting events. Several other recently described coarse-grained inclusions also have features consistent with a sequence of early, high-temperature melting, secondary alteration, and remelting at a lower temperature, suggesting that remelting of refractory inclusions was a common occurrence in the solar nebula.

Analyses of coarse-grained refractory inclusions typically do not have the solar CaO/Al2O3 ratio, probably reflecting nonrepresentative sampling of them in the laboratory. Many previous studies, especially those done by instrumental neutron activation analysis (INAA), were based on very small amounts of material removed from those restricted portions of inclusions that happened to be exposed on surfaces of bulk meteorite samples. Here, we address the sampling problem by studying thin sections of large inclusions, and by analyzing much larger aliquots of powders of these inclusions by INAA than has typically been done in the past. These results do show convergence toward the solar CaO/Al2O3 ratio of 0.792. The bulk compositions of 15 coarse-grained inclusions determined by INAA of samples >2 mg have an average CaO/Al2O3 ratio of 0.80 ± 0.18. When bulk compositions are obtained by modal recombination based on analysis of thin sections with cross-sections of entire, large, unbroken inclusions, the average of 11 samples (0.79 ± 0.15) also matches the solar value. Among those analyzed by INAA and by modal recombination, there were no inclusions for which both techniques agreed on a CaO/Al2O3 ratio deviating by >∼15% from the solar value. These results suggest that: individual inclusions may have the solar CaO/Al2O3 ratio; departures from this value are due to sample heterogeneity and nonrepresentative sampling in the laboratory; and it is therefore valid to correct compositions to this value. We present a method for doing so by mathematical addition or subtraction of melilite, spinel, or pyroxene. This yields a set of multiple, usually slightly different, corrected compositions for each inclusion. The best estimate of the bulk composition of an inclusion is the average of these corrected compositions, which simultaneously accounts for errors in sampling of all major phases. Results show that Type B2 inclusions tend to be more SiO2-rich and have higher normative Anorthite/Gehlenite component ratios than Type B1s. The inclusion bulk compositions lie in a field that can result from evaporation at 1700–2000K of CMAS liquids with solar CaO/Al2O3, but with a wide range of initial MgO (30–60 wt%) and SiO2 (15–50 wt%) contents.

On the night of March 26, 2003, a large meteorite broke up and fell upon the south suburbs of Chicago. The name Park Forest, for the village that is at the center of the strewnfield, has been approved by the nomenclature committee of the Meteoritical Society. Satellite data indicate that the bolide traveled from the southwest toward the northeast. The strewnfield has a southeast-northwest trend; however, this is probably due to the effects of strong westerly winds at high altitudes. Its very low 56Co and very high 60Co activities indicate that Park Forest had a preatmospheric mass that was at least ~900 kg and could have been as large as ~7 x 10^3 kg, of which only ~30 kg have been recovered. The average compositions of olivine and low-Ca pyroxene, Fa24.7 ± 1.1 and Fs20.8 ± 0.7, respectively, and its bulk oxygen isotopic composition, δ18O = +4.68‰, δ17O = +3.44‰, show that Park Forest is an L chondrite. The ferromagnesian minerals are well equilibrated, chondrules are easily recognized, and maskelynite is mostly ≤50 μm across. Based on these observations, we classify Park Forest as type 5. The meteorite has been strongly shocked, and based on the presence of maskelynite, mosaicism and planar deformation features in olivine, undulatory extinction in pyroxene, and glassy veins, the shock stage is S5. The meteorite is a monomict breccia, consisting of light-colored, angular to rounded clasts in a very dark host. The light and dark lithologies have essentially identical mineral and oxygen isotopic compositions. Their striking difference in appearance is due to the presence of a fine, pervasive network of sulfide veins in the dark lithology, resulting in very short optical path lengths. The dark lithology probably formed from the light lithology in an impact that formed a sulfide-rich melt and injected it into cracks.

Most studies of Tagish Lake have considered features that were either strongly affected by or formed during the extensive hydrous alteration experienced by this meteorite. This has led to some ambiguity as to whether Tagish Lake should be classified a CI, a CM, or something else. Unlike previous workers, we have focused upon the primary, anhydrous component of Tagish Lake, recovered through freeze-thaw disaggregation and density separation and located by thin section mapping. We found many features in common with CMs that are not observed in CIs. In addition to the presence of chondrules and refractory forsterite (which distinguish Tagish Lake from the CIs), we found hibonite-bearing refractory inclusions, spinel-rich inclusions, forsterite aggregates, Cr-, Al-rich spinel, and accretionary mantles on many clasts, which clearly establishes a strong link between Tagish Lake and the CM chondrites. The compositions of isolated olivine crystals in Tagish Lake are also like those found in CMs. We conclude that the anhydrous inclusion population of Tagish Lake was, originally, very much like that of the known CM chondrites and that the inclusions in Tagish Lake are heavily altered, more so than even those in Mighei, which are more heavily altered than those in Murchison.

Through freeze-thaw disaggregation of the Murchison (CM) carbonaceous chondrite, we have recovered a ~90 x 75 μm refractory inclusion that consists of corundum and hibonite with minor perovskite. Corundum occurs as small (~10 μm), rounded grains enclosed in hibonite laths (~10 μm wide and 30-40 μm long) throughout the inclusion. Perovskite predominantly occurs near the edge of the inclusion. The crystallization sequence inferred petrographically - corundum followed by hibonite followed by perovskite - is that predicted for the first phases to form by equilibrium condensation from a solar gas for Ptot ≤ 5 x 10^(-3) atm. In addition, the texture of the inclusion, with angular voids between subhedral hibonite laths and plates, is also consistent with formation of the inclusion by condensation. Hibonite has heavy rare earth element (REE) abundances of ~40 x CI chondrites, light REE abundances ~20 x CI chondrites, and negative Eu anomalies. The chondrite-normalized abundance patterns, especially one for a hibonite-perovskite spot, are quite similar to the patterns of calculated solid/gas partition coefficients for hibonite and perovskite at 10^(-3) atm and are not consistent with formation of the inclusion by closed-system fractional crystallization. In contrast with the features that are consistent with a condensation origin, there are problems with any model for the formation of this inclusion that includes a molten stage, relic grains, or volatilization. If thermodynamic models of equilibrium condensation are correct, then this inclusion formed at pressures

Berman’s (1983) activity-composition model for CaO-MgO-Al2O3-SiO2 liquids is used to calculate the change in bulk chemical and isotopic composition during simultaneous cooling, evaporation, and crystallization of droplets having the compositions of reasonable condensate precursors of Types A and B refractory inclusions in CV3 chondrites. The degree of evaporation of MgO and SiO2, calculated to be faithfully recorded in chemical and isotopic zoning of individual melilite crystals, is directly proportional to evaporation rate, which is a sensitive function of PH2, and inversely proportional to the droplet radius and cooling rate. When the precursors are partially melted in pure hydrogen at peak temperatures in the vicinity of the initial crystallization temperature of melilite, their bulk chemical compositions evolve into the composition fields of refractory inclusions, mass-fractionated isotopic compositions of Mg, Si, and O are produced that are in the range of the isotopic compositions of natural inclusions, and melilite zoning profiles result that are similar to those observed in real inclusions. For droplets of radius 0.25 cm evaporating at PH2 = 10−6 bar, precursors containing 8 to 13 wt.% MgO and 20 to 23% SiO2 evolve into objects similar to compact Type A inclusions at cooling rates of 2 to 12 K/h, depending on the precise starting composition. Precursors containing 13 to 14 wt.% MgO and 23 to 26% SiO2 evolve into objects with the characteristics of Type B1 inclusions at cooling rates of 1.5 to 3 K/h. The relatively SiO2-poor members of the Type B2 group can be produced from precursors containing 14 to 16 wt.% MgO and 27 to 33% SiO2 at cooling rates of 15% than are found on any condensation curve. The characteristics of fluffy Type A inclusions, including their reversely zoned melilite, can only be understood in the context of this model if they contain relict melilite.

Literature data for major element oxide compositions of most coarse-grained Types A and B inclusions in CV3 chondrites may be in error due to non-representative sampling of spinel relative to other phases because of small sample sizes. When reported compositions are corrected to the solar CaO/Al2O3 ratio by addition or subtraction of spinel, distinct trends result on oxide–oxide plots. These trends lie close to trajectories of bulk compositions of equilibrium condensates calculated for solar or dust-enriched gases under various conditions, except on a plot of MgO vs. SiO2 contents, where there is considerable scatter of the data points to the MgO-poor side of the condensation trajectory. The irreversible process of evaporative mass loss from a liquid droplet into an unsaturated H2 gas is modeled as a series of small equilibrium steps. This model is used to show that evolutionary paths of CMAS liquid compositions are identical for evaporation at all PH2 from 1 × 10−15 to 1 bar, with the ratio of the fraction of the SiO2 evaporated to that for MgO increasing both with increasing temperature from 1700 to 2000 K and with increasing SiO2 content of the starting composition. Such calculations show that compositions of most Type B inclusions can be explained by non-equilibrium evaporation of 10 to 30% of the MgO and 0 to 15% of the SiO2 into an H2 gas at 1700 K from liquid droplets whose compositions originated on any one of many possible equilibrium condensation trajectories. Some Type As may have suffered similar evaporative losses of MgO and SiO2 but at higher temperature. This degree of evaporation is consistent with the amount of Mg and Si isotopic mass fractionation observed in Types A and B inclusions. Evaporation probably happened after most Mg and Si were removed from the nebular gas into lower-temperature condensates.

— In addition to the Mg-, Al-, 16O-rich spinels that are known to occur in refractory inclusions, the Murchison meteorite contains Cr-rich, 16O-poor spinels, most of whose sources are unknown because they are rarely found in situ. Here we report the in situ occurrence in Allende of Cr-rich spinels, found in 13 chondrules and 4 “olivine-rich objects”. The Allende spinels exhibit major and minor element contents, isotopic compositions, and zoning of Cr2O3 contents like those of the Cr-spinels from Murchison. Some chondrules contain patchy-zoned spinel (Simon et al., 1994), which suggests that such grains did not form by sintering but perhaps by formation of overgrowths on relic grains. Unlike the olivine-rich objects, phases in all three chondrules that were analyzed by ion microprobe have uniform, near-normal O-isotopic compositions. One olivine-rich object, ALSP1, has a huge (1 mm) fragment of chevron-zoned spinel. This spinel has near-normal O-isotopic compositions that are quite distinct from those of adjacent forsteritic olivine, which are relatively 16O-rich and plot on the calcium-aluminum-inclusion (CAI) line, like some isolated forsterite grains found in Allende. The spinel and olivine in this object are therefore not genetically related to each other. Another olivine-rich object, ALSP11A, contains a rectangular, 150 ×s 100 μm, homogeneous spinel grain with 50 wt% Cr2O3 and 23 wt% FeO in a vuggy aggregate of finer-grained (5–90 μm), FeO-rich (Fo47–55) olivine. The magnesian core of one olivine grain has a somewhat 16O-rich isotopic composition like that of the large spinel, whereas the FeO-rich olivine is relatively 16O-poor. The composition of the spinel in ALSP11A plots on the CAI line, the first Cr-rich spinel found to do so. Chevron-zoned spinel has not been observed in chondrules, and it is unlikely that either ALSP1 or ALSP11A were ever molten. Calculations show that a spinel with the composition of that in ALSP1 can condense at 1780 K at a Ptot of 10−3 atm and a dust/gas ratio of 100 relative to solar. The Cr-rich spinel in ALSP11A could condense at ∼1420 K, but this would require a dust/gas enrichment of 1000 relative to solar. The data presented here confirm that, as in Murchison, the coarse Cr-rich spinels in Allende are relatively 16O-depleted and are isotopically distinct from the 16O-enriched MgAl2O4 from CAIs. Sample ALSP11A may represent a third population, one that is Cr-rich and plots on the CAI line. That the O-isotopic composition of ALSP1 is like those of Cr-rich spinels from chondrules indicates that O-isotopic compositions cannot be used to distinguish whether grains from such unequilibrated objects are condensates or are fragments from a previous generation of chondrules.

Compact type A (CTA) inclusions are one of the major types of coarse-grained refractory inclusions found in carbonaceous chondrites. They have not been studied in a systematic fashion, leading to some uncertainties and unproven assumptions about their origin. To address this situation, we studied a total of eight CTAs from Allende, Efremovka and Axtell by scanning electron-microscopic and electron and ion-microprobe techniques. These inclusions are very melilite-rich, ranging from ∼60 vol% to nearly monomineralic. Also present are Mg–Al spinel (5–20%), perovskite (trace–∼3%) and, in some samples, Ti-rich (∼17 wt% TiO2tot) fassaite (trace–∼20%), and rhonite (≤1%). Melilite compositions are mostly between Ak15 and Ak40. Chondrite-normalized REE abundance patterns for melilite (flat at ∼10 × CI with positive Eu anomalies) and fassaite (slight HREE enrichment relative to LREE and negative Eu anomalies) are like those for their counterparts in once-molten type B inclusions. The patterns for rhonite have positive slopes from La through Lu and abundances

— We have studied both of the known glass-free, hibonite-pyroxene spherules: MYSM3, from Murray (CM2), and Y17–6, from Yamato 791717 (CO3). They consist of hibonite plates (∼2 wt% TiOtot2) enclosed in Al-rich pyroxene that has such high amounts of CaTs (CaAl2SiO6) component, up to ∼80 mol%, that it must have crystallized metastably. Within the pyroxene, abundances of MgO and SiO2 are strongly correlated with each other and are anticorrelated with those of Al2O3, reflecting an anticorrelation between the diopside and CaTs components of the pyroxene. In contrast with previous results for Type B fassaite, however, we do not observe an anticorrelation between MgO and TiOtot2, possibly reflecting different relative distribution coefficients for Ti3+ and Ti4+ in the aluminous pyroxene of the spherules from those found for fassaite in Type B inclusions. Previously described hibonite-silicate spherules have 26Mg deficits but the present samples do not. Furthermore, the pyroxene in Y17-6 has excess 26Mg, while the hibonite it encloses does not, indicating that the two phases either had different initial 26Al/27Al ratios or different initial 26Mg/24Mg ratios. The Ti isotopic compositions of the present samples are highly unusual: δ50Ti = 103.4 ± 5.2%o in MYSM3 and -61.4 ± 4.1%0 in Y17-6, which are among the largest 50Ti anomalies reported for any refractory inclusion. The textures suggest that hibonite crystallized first; but based on the calculated bulk compositions of both spherules, it is not the liquidus phase in either sample, which suggests that the hibonite in both samples is relict. The presence of ragged hibonite grains in MYSM3 and rounded hibonite grains in Y17-6 and a lack of isotopic equilibrium between pyroxene and hibonite support this conclusion. The spherules crystallized from liquid droplets that probably formed as a result of the melting of solid precursor grains that included hibonite. The heating events were too short and/or not hot enough to melt all the hibonite. The droplets cooled quickly enough that CaTs-rich pyroxene crystallized instead of anorthite. Based on the observed differences in isotopic composition, it is unlikely that the precursors of the present samples formed in the same reservoir as each other or as the previously described hibonite-silicate spherules, providing further evidence of the isotopic heterogeneity of the early solar nebula.

— We report the results of a study of TS2, an unusual compact Type A inclusion from Allende. A distinctive, major feature of this inclusion is that many of its melilite crystals have no dominant core-rim zoning but instead consist of 50–200 μm patches of Mg-rich melilite (Ak32–62, median Ak51) set in or partially enclosed by, and optically continuous with, relatively Al-rich melilite (Ak25–53, median Ak38). The Al-rich regions have jagged, dendritic shapes but occur within crystals having straight grain boundaries. Another unusual feature of this inclusion is the size and spatial distribution of spinel. In many places, especially in the interior of the inclusion, the aluminous melilite encloses numerous, fine (0.5–5 μm) inclusions of spinel and minor perovskite and fassaite. The latter phases also occur as isolated grains throughout the inclusion. Coarse-grained spinel, ∼50–150 μm across, occurs in clumps and chains enclosed in relatively Mg-rich melilite, whereas none of the fine spinel grains are clumped together. The sample also contains a spinel-free palisade body, 1.7 × 0.85 mm, that consists almost entirely of Ak-rich (45–65 mol%) melilite. Within the palisade body are two grains of perovskite with extremely Nb-rich (∼4–8 wt% Nb2O5) cores and rims of typical composition. All phases in this inclusion have chondrite-normalized REE patterns that are consistent with crystal/melt partitioning superimposed upon a bulk modified Group II pattern. We suggest that TS2 had an anomalous cooling history and favor the following model for the formation of TS2. Precursors having a bulk modified Group II pattern melted. Rapid growth of large, dendritic, nonstoichiometric melilite crystals occurred. The melilite trapped pockets of melt and incorporated excess spinel components and TiO2. Bubbles formed in the residual melt. As crystallization slowed, coarse spinel grew. Some spinel grains collected against bubbles, forming spherical shells, and others formed clumps and chains. Relatively Ak-rich melilite crystallized from the residual melt between dendritic melilite crystals and from melt trapped in pockets and between arms of dendrites, and incorporated the clumps and chains of coarse spinel. Bubbles broke and filled with late-stage melt, their shapes preserved by their spinel shells. Slow cooling, or perhaps an episode of reheating, allowed the early melilite to become stoichiometric by exsolving fine grains of spinel, perovskite and fassaite, and allowed the melilite to form smooth grain boundaries. Dendritic crystals are indicative of rapid growth and the melilite crystals in TS2 appear to be dendritic. Coarse, dendritic melilite crystals have been grown from Type B inclusion melts cooled at ∼50–100 °C/h. If those results are applicable to Type A inclusions, we can make the first estimate of the cooling rate of a Type A inclusion, and it is outside the range (2–50 °C/h) generally inferred for Type B inclusions. The rapid cooling inferred here may be part of an anomalous thermal history for TS2, or it may be representative of part of a normal thermal history common to Types A and B that involved rapid cooling early (at high temperatures) as inferred for TS2, and slower cooling later (at lower temperatures), as inferred for Type B inclusions. We prefer the former explanation; otherwise, the unusual features of TS2 that are reported here would be common in Type A inclusions (which they are not).

— Through freeze-thaw disaggregation of the Murchison meteorite, we have recovered, in addition to many spinel-hibonite spherules, several hibonite-rich inclusion fragments in which the hibonite has wider ranges in TiO2 contents (e.g., 0.07–8.6 wt% in one inclusion and 2–10 wt% in another) than previously observed within single inclusions. In these inclusions, there are sharp contacts between texturally early, Ti-poor hibonite and relatively late, Ti-rich hibonite, and the two types occur in complex intergrowth textures that are not consistent with crystallization from a melt in a single-stage cooling event. One inclusion has, in addition to relatively TiO2-rich hibonite, some that is virtually TiO2-free but contains ∼1 wt% MgO and ∼1.5 wt% SiO2. Instead of the common substitution of Mg + Ti for 2Al, Mg coupled with Si in this case, probably reflecting crystallization from an unusual, Ti-free silicate liquid. Ion microprobe analyses of Ti-rich and Ti-poor hibonite from this inclusion yield quite similar trace-element patterns and Mg-isotopic compositions. The results are most consistent with formation of Ti-rich hibonite from Ti-poor hibonite by addition of Mg and Ti to the latter by exchange with a hot, Ti-rich liquid. That this occurred without a resolvable change in the Mg-isotopic composition requires that the Ti-rich, second generation of hibonite formed

— It has been suggested that palisade bodies—shells of spinel found within some calcium, aluminum-rich inclusions (CAIs) and the phases the shells enclose—are intact mini-CAIs that predate and were captured by their current hosts while the latter were still molten. We present new data and observations that indicate that most palisade bodies formed instead in situ while their host inclusions were crystallizing. The evidence includes observations of spinel-lined cavities and glass-filled, circular structures outlined by spinel in experimental run products crystallized from melts; a partially formed palisade body in an inclusion; a fassaite crystal that is optically continuous across a palisade wall; and similarity of unusual mineral compositions in some palisade bodies and their hosts. Our observations can be used to refute arguments for exotic origin and are most consistent with a model for in situ formation involving: (1) formation of vesicles in a largely molten inclusion; (2) nucleation of spinel upon and/or adherence to vapor-melt interfaces, forming spinel shells around vesicles; (3)leakage of vesicles and filling with melt while spinel shells remain largely intact; and (4) crystallization of melt inside shells. This model is similar to one proposed for formation of segregation vesicles, which are partially- to completely-filled vesicles found in some terrestrial basalts. In addition, we interpret framboids (i.e., dense clusters of spinel with little material between grains, found in most inclusions that contain palisade bodies) as polar or near-polar sections through palisade bodies and therefore do not make a genetic distinction between the two features.

— Through freeze-thaw disaggregation of the Murchison meteorite, we have recovered a refractory inclusion, HIB-11, that is unique in terms of its texture, mineral compositions, and bulk composition. It consists of anhedral, Y-rich (1.6 wt% Y2O3) perovskite and lathlike spinel grains enclosed in a matrix of fine-grained, Sc-rich (10.5 wt% SC2O3 avg.), Ti-rich (12.6 wt% TiO2 avg., reporting all Ti as TiO2) clinopyroxene. The chondrite-normalized rare earth element (REE) pattern is complex, with light REE (LREE) at ∼10× C1, abundances increasing from Gd through Ho (the latter at ∼104× C1), decreasing through Yb at 200× C1, and Lu at ∼400× C1. The pattern reflects several stages of high-temperature volatility fractionation. Removal of Lu and Er from the source gas in the first condensation event was followed by partial to complete removal of the somewhat less refractory heavy REE, Gd through Ho, in the HIB-11 precursors by condensation from the fractionated residual gas in a second event. Both of these events probably reflect condensation of REE into ZrO2 or a mixed Zr-, Sc-, Ti-, Y-oxide at temperatures too high for hibonite stability. A second, lower-temperature component, which was subsequently added, had fractionated (Nd-poor, Ce-rich) LREE abundances that resulted from condensation from a gas that had undergone prior removal of the more refractory LREE, resulting in enrichment in Ce and the most volatile REE, Eu and Yb. The aggregate was then melted and quickly cooled, forming a fine-grained spherule. This is the first reported inclusion in which the two most refractory REE, Lu and Er, are strongly fractionated from the other REE. An absence of mass fractionation among the Ti isotopes indicates that HIB-11 is not an evaporative residue, implying that volatility fractionation of trace elements took place during condensation. The fact that the two most refractory heavy REE could be separated from the other, only slightly less refractory heavy REE suggests that a wide variety of REE patterns is possible, and that ultrarefractory inclusions with other unusual REE patterns, important recorders of nebular condensation, may yet be discovered.

The Berman (1983) activity-composition model for non-ideal liquid solutions in the CaO MgO Al2O3 SiO2 (CMAS) system is incorporated into equilibrium condensation calculations which are used to explore the vapor-solid-liquid stability relations as a function of temperature and total pressure (Ptot) in a gas of solar composition, and as a function of temperature and dust/gas ratio at Ptot = 1 × 10−3 atm in gases produced by total vaporization of systems enriched in interstellar dust relative to the gas compared to solar abundances. Condensate liquids are very non-ideal, suggesting that results of previous attempts to model their formation using ideal solution models are highly inaccurate. As is the case for the Berman (1983) liquid model itself, results of the present calculations are in very good agreement with experimentally determined liquid-crystal phase relations except where intermediate members of solid solution series, such as melilite and fassaite, are predicted to be in equilibrium with liquid, in which cases liquid-crystal equilibration temperatures are overestimated by 50 to 100 K. CMAS liquids are stable in a solar gas at a Ptot at least as low as 5 × 10−2 atm and perhaps as low as 1 × 10−2 atnt, much lower than previous estimates for liquids of pure forsterite composition, due to the colligative effects of CaO and Al203. CMAS liquids are stable at Ptot = 1 × 10−3 atm in systems with dust/gas enrichment factors at least as low as 16 and perhaps as low as 5 relative to solar abundances. Results of these calculations suggest that, upon cooling, a solid melilite + spinel condensate assemblage, comparable to a Type A refractory inclusion, would react with the vapor to produce a liquid much richer in MgO and SiO2 than the starting material, at either elevated Ptot or enhanced dust/gas ratio. If this partial melt were isolated from further reaction with the nebular gas, it would solidify into a spinel + melilite + fassaite + anorthite assemblage, similar in chemical and mineralogical composition to a Type B refractory inclusion. Forsterite coexists stably with CMAS condensate liquids over wide ranges of Ptot and dust/gas ratio, extending to the lowest Ptot and dust/gas ratio at which liquids are stable. If the compositions of glass inclusions inside isolated forsterite crystals in the Murchison CM2 chondrite have been modified by precipitation of 25 wt% forsterite as a daughter mineral from the liquid precursors of those glasses, the inclusions could represent condensate liquids that were in equilibrium with forsterite at Ptot = 0.3 atm or at dust/gas enrichment factors of ∼70 at Ptot = 1 × 10−3 atm.

Watson, which was found in 1972 in South Australia, contains the largest single silicate rock mass seen in any known iron meteorite. A comprehensive study has been completed on this unusual meteorite: petrography, metallography, analyses of the silicate inclusion (whole rock chemical analysis, INAA, RNAA, noble gases, and oxygen isotope analysis) and mineral compositions (by electron microprobe and ion microprobe). The whole rock has a composition of an H-chondrite minus the normal H-group metal and troilite content. The oxygen isotope composition is that of the silicates in the IIE iron meteorites and lies along an oxygen isotope fractionation line with the H-group chondrites. Trace elements in the metal confirm Watson is a new IIE iron. Whole rock Watson silicate shows an enrichment in K and P (each approximately 2X H-chondrites). The silicate inclusion has a highly equilibrated igneous (peridotite-like) texture with olivine largely poikilitic within low-Ca pyroxene: olivine (Fa20), opx (Fs17Wo3), capx (Fs9Wo14)(with very fine exsolution lamellae), antiperthite feldspar (An1-3Or5) with less than 1 micron exsolution lamellae (An1-3Or greater than 40), shocked feldspar with altered stoichiometry, minor whitlockite (also a poorly characterized interstitial phosphate-rich phase) and chromite, and only traces of metal and troilite. The individual silicate minerals have normal chondritic REE patterns, but whitlockite has a remarkable REE pattern. It is very enriched in light REE (La is 720X C1, and Lu is 90X C1, as opposed to usual chonditic values of approximately 300X and 100-150X, respectively) with a negative Eu anomaly. The enrichment of whole rock K is expressed both in an unusually high mean modal Or content of the feldspar, Or13, and in the presence of antiperthite.

Most spinel grains in Murchison acid residues are Mg-, Al-rich, 16O-rich (δ18O = −50%.), small (10–30 μm) and probably from refractory inclusions. They are quite unlike spinels we have recovered from Murchison by freeze-thaw disaggregation, density separation, and handpicking. As reported here, the latter spinels contain up to 37 wt% Cr2O3 and up to 17 wt% FeO, are not 16O-enriched (δ18O = 1.9 ± 2.4%.), are coarse (60–325 μm), and are not from refractory inclusions. From backscattered electron images of fifty-seven such grains, we recognize five zoning types defined by variations in Cr2O3 contents: patchy (56%); homogeneous (21%); chevron (10.5%); gradational (9%); and core-rim (3.5%). Many grains have silicate inclusions, the most common being small, anhedral grains of diopside with 12–24 wt% Al2O3 and up to 3.7 wt% TiO2. Eleven spinel samples occur with forsteritic (Fo95–99) olivine; in most cases, the spinel partially encloses the olivine. Cr-bearing spinel was found in situ in two Al-rich chondrules (one with homogeneous spinel, the other with homogeneous, gradational and core-rim spinel, and both with forsterite and aluminous diopside); in two irregularly shaped, olivine-bearing inclusions (one with homogeneous spinel, the other patchy); and attached to an isolated olivine grain (patchy). Observation of homogeneous, gradational, and core-rim type spinels in chondrules and basalts shows that grains with these zoning patterns can crystallize from liquids, although, in Murchison, chondrules with the appropriate compositions and sufficiently coarse textures to yield the separated spinels are exceedingly rare. Chevronzoned grains also could have formed in chondrules; alternatively, they may have acquired their oscillatory zoning patterns by cycling through different P-T-ƒ O 2 regimes in the solar nebula during their formation. The patchy spinel grains were probably never molten and they most likely formed by sintering of aggregates of smaller spinel grains which were enriched in Cr and Fe to varying degrees. In spite of their various crystallization and thermal histories, the spinels all have normal oxygen and chromium isotopic compositions, consistent with formation from a single, well-mixed nebular reservoir. Based on the known slow rates of diffusion of oxygen in MgO, Al2O3, and MgAl2O4, it is unlikely that the spinels of this study formed from an isotopically anomalous reservoir and later re-equilibrated with a normal one; it is more likely that they have retained their original isotopic compositions. We see no evidence for anomalous Cr, which had been reported by others.

– Equilibrium thermodynamic calculations of the sequence of condensation of phases from a cooling gas of solar composition at total pressures thought to have prevailed in the inner part of the solar nebula successfully predict the primary mineral assemblages of refractory inclusions in CM2 and CV3 chondrites. Many refractory inclusions in CM2 chondrites contain a relatively SiO2-poor assemblage (spinel, hibonite, grossite, perovskite, corundum) that represents a high-temperature stage of condensation, and some may be pristine condensates that escaped later melting. Compact Type A and Type B refractory inclusions, consisting of spinel, melilite, perovskite, Ca-rich clinopyroxene ± anorthite, in CV3 chondrites are more SiO2-rich and equilibrated with the solar nebular gas at a slightly lower temperature. Textures of many of these objects indicate that they underwent melting after condensation, crystallizing into the same phase assemblage as their precursors. The Ti3+/Ti4+ ratio of their pyroxene indicates that this process occurred in a gas whose oxygen fugacity () was approximately 8.5 log units below that of the iron-wustite buffer, making them the only objects in chondrites known to have formed in a system whose composition was close to that of the sun. Relative to CI chondrites, these inclusions are uniformly enriched in a group of elements (e.g., Ca, REE, Zr, Ta, Ir) that are chemically diverse except for their high condensation temperatures in a system of solar composition. The enrichment factor, 17.5, can be interpreted to mean that these objects represent either the first 5.7 wt% of the condensable matter to condense during nebular cooling or the residue after vaporization of 94.3% of a CI chondrite precursor. The Mg and Si isotopic compositions of Types A and B inclusions are mass-fractionated by up to 10 and 4 ‰/amu, respectively. When interpreted in terms of Rayleigh fractionation during evaporation of Mg and Si from the inclusions while they were molten, the isotopic compositions imply that up to 60% of the Mg and up to 25% of the Si were evaporated, and that approximately 80% of the enrichment in refractory (CaO+Al2O3) relative to more volatile (MgO+SiO2) in the average inclusion is due to initial condensation and approximately 20% due to subsequent evaporation. The mineralogical composition, including the Ti3+/Ti4+ ratio of the pyroxene, in Inti, a particle sampled from Comet Wild 2 by the Stardust spacecraft, is nearly identical to that of a Type B inclusion, indicating that comets contain not only the lowest-temperature condensates in the form of ices but the highest-temperature condensates as well. The FeO/(FeO+MgO) ratios of olivine and pyroxene in the matrix and chondrules of carbonaceous and ordinary chondrites are too high to be made in a system of solar composition, requiring s only 1 or 2 log units below iron-wustite, more than 105 times higher than that of a solar gas. Various ways have been devised to generate cosmic gases sufficiently oxidizing to stabilize significant FeO in olivine at temperatures above those where Fe-Mg interdiffusion in olivine ceases. One is by vertical settling of dust toward the nebular midplane, enriching a region in dust relative to gas. Because dust is enriched in oxygen compared to carbon and hydrogen relative to solar composition, a higher results from total vaporization of the region, but the factor by which theoretical models have so far enriched the dust is 10 times too low. Another is by transporting icy bodies from the outer part of the nebula into the hot, inner part where vaporization of water ice occurs. Not only does this method fail to make the needed by a factor of 30–1000 but it also ignores simultaneous evaporation of carbon-bearing ices that would make the even lower.

The discovery is reported of opaque assemblages in the Leoville carbonaceous chondrite with both clear textural and chemical evidence of a major role for resolution and equilibration at relatively low temperatures in the formation of refractory siderophile element-rich assemblages. The observed phase compositions are used to bracket a previously inferred phase boundary in the Fe-Ni-Ru system. The first meteoritic occurrence of goldmanite is also reported.

Major-, minor-, and trace-element zoning trends in fassaite from Type B1 and B2 inclusions from the Allende meteorite are defined by using electron and ion microprobe data. Fassaite/liquid distribution coefficients are calculated for many of the lithophile elements, and evidence for and against the presence of relict fassaite in these inclusions is discussed. Major-element zoning in fassaite in Type B CAIs is found from Ti-, V-, Sc-rich cores to relatively Mg-, Si-rich rims. Trace elements that are incompatible in this fassaite are Nb (effective fassaite/liquid distribution = 0.29), Y(0.52), Ta(0.34), Th(0.053), U(0.4), and rare earth elements (0.31-0.48), while compatible trace elements are Zr(1.1) and Hf(1.5). Rare-earth-elements (RRE) abundances in fassaite are found to be controlled by crystal/liquid fractionation. Candidates for relict fassaite grains are discovered: Ti-, Sc-rich, REE-poor crystals enclosed in melilite in the mantle of TS34.

Results are presented on a detailed petrologic, chemical, and isotopic study of the so-called FUN inclusion (1623-5) from the Vigarano C3V chondrite. It is shown that the precursor material from which the Vigarano 1623-5 has formed contained some nuclear isotopic anomalies; this precursor was composed of melted and crystallized spinel, olivine, fassaite, and melilite. The results on the petrologic and isotopic properties of 1623-5 indicate unambiguously the action of volatilization in the evolution of this inclusion.

著者人数: 75名, 資料番号: SA1003700000

Inclusions called 'fluffy' Type A's or FTA's in the Allende meteorite are examined as possible candidates for relict vapor-solid condensate grains, remaining from the original solar nebula. Type A inclusions are characterized by abundant melilite and absence of primary anorthite and titaniferous pyroxene. Fluffy Type A's were probably loosely bound clumps of crystals drifting in the solar nebula, analogous to dustballs or snowflakes. Polished thin sections of all samples were studied optically and with a JEOL JSM-35 scanning electron microscope. It is reasonably clear that neither whole FTA's nor constituent nodules of the coarser grained ones were ever molten. Despite solid-state recrystallization which has affected these inclusions to varying degrees, the coarser grained material remaining in many of them is probably a relic of vapor-solid condensation in the solar nebula.

The use of the ion microprobe for quantitative analysis of Sr, Y, Zr, La, Sm, and Yb in melilite and pyroxene is evaluated. Three trace element-doped synthetic glasses of composition Ak_40, Ak_80, and Di_2AbAn were analyzed by ion microprobe (IMP) using ion yields determined from Corning glass standards. IMP-determined oxide concentrations in the Di_2AbAn glass agree well with electron microprobe (EMP) analyses (to within 6%), but IMP analyses of the melilite glasses deviate from EMP averages by up to 19%. The deviations are due to erroneous SiO_2 estimates caused by suppression of Si ion intensities by the enhanced concentrations of Ca and Al in the melilite glasses compared to the standards. Thus, in order to determine compositions of melilite, diopside, and glass from subliquidus experiments on each of the three starting compositions, we adopted a new set of ion yields such that IMP analyses of the three starting glasses reproduce the EMP average compositions. Further IMP and EMP comparisons of the subliquidus assemblages show that quantitative analyses of melilite, diopside, and glass can be obtained by IMP that are within 10% of the concentrations obtained by EMP, when ion yields determined from glass starting compositions are used. EMP-IMP comparison of crystal and glass analyses also suggests that a structural matrix effect may result in overestimation of SrO (10–12%) in melilite by IMP. Comparison of our data for Ak_12 and Ak_90 melilite compositions with literature results shows that melilite/liquid for REE^3+ determined by IMP decrease with increasing X_Ak (Ak_90: D_La = 0.038, D_Sm = 0.032, D_Yb = 0.0086; Ak: 0.67, 0.75, 0.25, respectively) while that for Sr (=Eu^2+) changes only slightly (0.99 to 0.78, respectively). Since X_Ak increases with decreasing temperature for all melilite with X_Ak

A detailed mineralogical study of HAL was initiated to elucidate the origin of this inclusion because LEE et al. (1979) had found large mass fractionation effects and small nuclear effects in its Ca isotopic composition, but no ^(26)Mg excesses in samples of it with very high ^(27)Al/^(24)Mg ratios. HAL's 1.0 mm radius interior consists almost entirely of three hibonite crystals and is surrounded by a 2.0 mm thick, multi-layered rim. The first layer, called the black rim, is black and compact, resembles a devitrified glass and contains an anisotropic Al-Fe-oxide similar to hercynite in composition. This is followed by a friable rim sequence, layer I of which is predominantly feldspathoids with minor anorthite, Ti-Fe-oxide and Al-Fe-oxide. Layer II contains abundant perovskite, plus grossular, andradite and pyroxene in addition to the minerals of layer I. Layer III is mostly Ca-phosphate, possibly hydroxy-apatite and perovskite. Layer IV is rich in hibonite, Al-Fe-oxide, perovskite, nepheline and the two garnets, lacks Ca-phosphate but contains traces of a Ti-Sc-Zr-oxide. Layer V is rich in Al-Fe-oxide, pyroxene, nepheline, the two garnets and olivine whose crystals display peculiar rectangular cross-sections. The black rim does not completely surround the hibonite core. Sectors of the friable rim exist where layer I is missing and where the mineralogy of adjacent layers is no different from that of the same layers in other sectors. Pentlandite, nickel-iron and barrel-shaped olivine crystals, minerals typical of the matrix of Allende and found nowhere else in HAL, are found in layer V and increase in abundance toward its exterior, as if grains of these phases accreted together with the other minerals of layer V. This layer also contains alternating olivine-rich and garnet-, pyroxene-rich bands, resembling rhythmic layering. For these reasons, we conclude that each of the layers of the friable rim formed by the accretion of an assemblage of condensate grains rather than by complete reaction of a precursor to HAL with a nebular gas. Thus, the unusual isotopic characteristics of HAL are thought to have been inherited from a nebular reservoir which was isotopically distinct from that which gave rise to the bulk of the material in Allende. HAL's mineralogical peculiarities indicate that its formation reservoir was also chemically distinct from the latter one.

Greyish-brown irregularly-shaped aggregates composed predominantly of olivine make up nearly 2% of the Allende meteorite by volume. Many of the aggregates are constructed of subspherical lumps of micron-sized crystals of olivine, pyroxene, nepheline and sodalite surrounded by coarser-grained olivine. Rarely, anorthite, spinel and perovskite are also present. The olivine ranges in composition from Fo64 to Fo99. Pyroxenes range from aluminous diopside to hedenbergite to very Al-rich and Ti-Al-rich varieties. The nepheline contains 1.6-2.4% K2O and 1.6-5.2% CaO but the sodalite is significantly poorer in these elements. The spinel contains 2.1-13.4% FeO. Textural information and oxygen isotopic data suggest that the aggregates are composed of primary, solid condensates from the solar nebula. The perovskite, spinel and Ti-Al-rich pyroxenes are the remains of high-temperature condensates, but the olivine compositions and the presence of feldspathoids indicate that some of the grains continued to react with the solar nebular vapor in the temperature range 500-900 K.

Electron spin resonance has been used to make the first direct determination of Ti^(3+) in synthetic hibonite and hibonite from inclusion SH-7 of the Murchison C2 chondrite. Ti^(3+) concentrations range from 0.02 to 0.64 wt% in synthetic blue hibonite and 0.35–0.44 wt% in hibonite from SH-7. No Ti^(3+) could be detected in orange hibonite, supporting the earlier conclusion that the orange-to-blue transition is associated with the presence of Ti^(3+). At constant temperature and oxygen fugacity, Ti^(3+)/Ti^(4+) in synthetic hibonite increases with decreasing V but is not strongly dependent on bulk Ti. At the concentration levels encountered in meteoritic hibonite, Fe and Cr contents do not have a significant effect on the amount of Ti^(3+). In both synthetic and meteoritic hibonite, Ti^(3+) occupies a 5-coordinated crystallographic site, which is consistent with the formation of doubly ionized oxygen vacancies. At low oxygen fugacities, essentially all Ti^(4+) on the five-fold Al-site has been reduced to Ti^(3+). Hibonite from SH-7 equilibrated with a gas that could have been as reducing as a gas of solar composition. This is consistent with other estimates based on mineral equilibria of high temperature oxygen fugacities in Ca-Al-rich inclusions. With the possible exception of Mo-W depletions, indicators based on bulk trace element concentrations in CAIs are inconclusive. There is considerable evidence that as CAIs cooled to lower temperatures, they experienced conditions significantly more oxidizing than those of a solar gas, perhaps in planetary environments.

Enthalpies of formation were determined for two calcium aluminate phases, CaAl4O7 and CaAl12O19, using high-temperature alkali borate solution calorimetry. The aluminates were synthesized by multiple-cycle heating and grinding stoichiometric mixtures of CaCO3 and Al2O3, and the products were characteized by X-ray diffraction and SEM microbeam analysis. The data on impurities (CaAl4O7 was found to be about 89.00 percent pure by weight and the CaAl12O19 samples about 91.48 percent pure) were used to correct the heat of solution values of the synthetic products. The enthalpies of formation, at 1063 K, from oxides, were found to be equal to -(25.6 + or - 4.7) kJ/g.f.w. for CaAl4O7 and -(33.0 + or - 9.7) kJ/g.f.w. for CaAl12O19; the respective standard enthalpies of formation from elements, at 298 K, were estimated to be -4007 + or - 5.2 kJ/g.f.w. and -10,722 + or - 12 kJ/g.f.w.

Isotopic and chemical analyses of hibonite-rich inclusions from a number of primitive meteorites and, in some cases, of coexisting minerals from carbonaceous and ordinary chondrites, were carried out using ion microprobe (IP). The results of the IP analyses were compared with literature data obtained by INAA and electron microprobe analysis. Results showed that the trace element patterns of the hibonite inclusions analyzed have a number of feautures in common. The results are discussed with reference to the possible chemical processes operating early in the history of the solar system.

The black inclusion from the Krymka LL3 chondrite previously found to contain 'mysterite' by Lewis et al. (1979) belongs to a hitherto unknown class of carbonaceous chondrites. Its olivine and pyroxene compositions, Fo 97-99 and En 96, respectively, are characteristic of carbonaceous chondrites and its plagioclase composition. An 100 is characteristic of C3's. It contains a peculiar group of Co-, Cr-rich metal grains whose compositions are similar, but not identical, to those in C2 chondrites. Its weight ratios of total Fe/SiO2 and SiO2/MgO are 0.74 and 1.43, respectively, and its atomic ratio of Si/Al is 10.7, exactly the same as in carbonaceous chondrites. Its bulk chemical composition is very close to that of the Murchison C2 chondrite. The association of mysterite with a special type of carbonaceous chondrite material suggests that mysterite formed by low-temperature condensation in a different region of the nebula from other carbonaceous chondrites.

There are two types of white, coarse-grained, Ca-Al-rich inclusions in Allende. Type A inclusions contain 80–85 per cent melilite, 15–20 per cent spinel, 1–2 per cent perovskite and rare plagioclase, hibonite, wollastonite and grossularite. Clinopyroxene, if present, is restricted to thin rims around inclusions or cavities in their interiors. Type B inclusions contain 35–60 per cent pyroxene, 15–30 per cent spinel, 5–25 per cent plagioclase and 5–20 per cent melilite. The coarse pyroxene crystals in Type B's contain >15 per cent Al2O3 and >1.8 per cent Ti, some of which is trivalent. Type A pyroxenes contain Electron microprobe analyses of 600 melilite, 39 pyroxene, 35 plagioelase, 33 spinel and 20 perovskite grains were performed in 16 Type A, 1 intermediate and 9 Type B inclusions in Allende and 1 Type A in Grosnaja. Melilite composition histograms from individual Type A inclusions are usually peaked between Ak10 and Ak30 and are 15–20 mole % wide while those from Type B inclusions are broader, unpeaked and displaced to higher akermanite contents. Most pyroxenes contain Inferred bulk chemical compositions of Type A inclusions are rather close to those expected for high-temperature condensates. Those of Type B inclusions suggest slightly lower temperatures but their Ca/Al ratio seems less than the Type A's, indicating that the Type B's may not be their direct descendants. Some textural features suggest that the inclusions are primordial solid condensetes while others indicate that they may have been melted after condensation. Fragmentation and metamorphism may have also occurred after condensation.

The primary phases of Al-rich inclusions in amoeboid olivine aggregates have undergone alteration reactions with the solar nebular gas. The simplest interpretation of the present observations is that melilite was the first primary phase to disappear with falling temperature, and was replaced by grossular + anorthite + feldspathoids, followed by fassaite; spinel was the last phase to be altered. Thermodynamic calculations suggest that Na-rich phlogopite could have formed at about 470 K and chlorite at about 328 K at a water fugacity of 0.000001, which is that of a gas of solar composition in this temperature range. The olivine around Al-rich inclusions is not serpentized, indicating the cessation of gas-solid equilibrium above 274 K.

Although thermodynamic calculations predict corundum to be the first condensate from a cooling gas of solar composition, a corundum-hibonite inclusion, BB-5, has for the first time been found in the Murchison carbonaceous chondrite. Ion microprobe measurements of Mg isotopic compositions yield the result, unexpected in such an early condensate, that Mg-26 excesses are small despite large Al-27/Mg-24 ratios. The extreme temperature required to melt this inclusion makes a liquid origin unlikely, except by the hypervelocity impact of refractory bodies. Alternatively, B-5 is a direct gas-solid condensate, and its uniform Mg-26 enrichment must be a characteristic of the reservoir from which it condensed. Nebular heterogeneity in magnesium isotopic composition is the preferred explanation for the formation of such a reservoir.