Non-spherical lobate chondrules in CO3.0 Y-81020: General implications for the formation of low-FeO porphyritic chondrules in CO chondrites

Non-spherical chondrules (arbitrarily defined as having aspect ratios ≥1.20) in CO3.0 chondrites comprise multi-lobate, distended, and highly irregular objects with rounded margins; they constitute ∼70% of the type-I (low-FeO) porphyritic chondrules in Y-81020, ∼75% of such chondrules in ALHA77307, and ∼60% of those in Colony. Although the proportion of non-spherical type-I chondrules in LL3.0 Semarkona is comparable (∼60%), multi-lobate OC porphyritic chondrules (with lobe heights equivalent to a significant fraction of the mean chondrule diameter) are rare. If the non-spherical type-I chondrules in CO chondrites had formed from totally molten droplets, calculations indicate that they would have collapsed into spheres within ∼10⁻³ s, too little time for their 20-μm-size olivine phenocrysts to have grown from the melt. These olivine grains must therefore be relicts from an earlier chondrule generation; the final heating episode experienced by the non-spherical chondrules involved only minor amounts of melting and crystallization. The immediate precursors of the individual non-spherical chondrules may have been irregularly shaped chondrule fragments whose fracture surfaces were rounded during melting. Because non-spherical chondrules and “circular” chondrules form a continuum in shape and have similar grain sizes, mineral and mesostasis compositions, and modal abundances of non-opaque phases, they must have formed by related processes. We conclude that a large majority of low-FeO chondrules in CO3 chondrites experienced a late, low-degree melting event. Previous studies have shown that essentially all type-II (high-FeO) porphyritic chondrules in Y-81020 formed by repeated episodes of low-degree melting. It thus appears that the type-I and type-II porphyritic chondrules in Y-81020 (and, presumably, all CO3 chondrites) experienced analogous formation histories. Because these two types constitute ∼95% of all CO chondrules, it is clear that chondrule recycling was the rule in the CO chondrule-formation region and that most melting events produced only low degrees of melting. The rarity of significantly non-spherical, multi-lobate chondrules in Semarkona may reflect more-intense heating of chondrule precursors in the ordinary-chondrite region of the solar nebula.     

The formation of chondrules by open-system melting of nebular condensates

Experiments were conducted under canonical nebular conditions to see whether the chemical compositions of the various chondrule types can be derived from a single CI-like starting material by open-system melting and evaporation. Experimental charges, produced at 1580 °C and P_H2 of 1.31×10⁻⁵ atm over 1 to 18 hours, consisted of only two phases, porphyritic olivine crystals in glass. Sulfur, metallic-iron and alkalis were completely evaporated in the first minutes of the experiments and subsequently the main evaporating liquid oxides were FeO and SiO₂. Olivines from short runs (2-4 hours) have compositions of Fo₈₃-Fo₈₉, as in Type IIA chondrules, while longer experimental runs (12-18 hours) produce ∼Fo₉₉ olivine, similar to Type IA chondrules. The concentration of CaO in both olivine (up to 0.6 wt.%) and glass, and their Mg#, increased with increasing heating duration. Natural chondrules also show increasing CaO with decreasing S, alkalis, FeO and SiO₂. The similarities in bulk chemistry, mineralogy and textures between Type IIA and IA chondrules and the experimental charges demonstrate that these chondrules could have formed by the evaporation of CI precursors. The formation of silica-rich chondrules (IIB and IB) by evaporation requires a more pyroxene-rich precursor.Based on the FeO evaporation rates measured here, Type IIA and IA chondrules, were heated for at least ∼0.5 and ∼3.5 h, respectively, if formed at 1580 °C and P_H2 of 1.31×10⁻⁵ atm. Type II chondrules may have experienced higher cooling-rates and less evaporation than Type I.The experimental charges experienced free evaporation and exhibited heavy isotopic enrichments in silicon, as well as zero concentrations of S, Na and K, which are not observed in natural chondrules. However, experiments on potassium-rich melts at the same pressure but in closed capsules showed less evaporation of K, and less K isotopic mass fractionation, than expected as a function of decreasing cooling rate. Thus the environment in which chondrules formed is as important as the kinetic processes they experienced. If chondrule formation occurred under conditions in which evaporated gases remained in the vicinity of the residual melts, the extent of evaporation would be reduced and back reaction between the gas and the melt could contribute to the suppression of isotopic mass fractionation. Hence chondrule formation could have involved evaporative loss without Rayleigh fractionation. Volatile-rich Type II and volatile-poor Type I chondrules may have formed in domains with high and low chondrule concentrations, and high partial pressures of lithophile elements, respectively.     

Sulfur isotope composition of putative primary troilite in chondrules from Bishunpur and Semarkona

The sulfur isotopic compositions of putative primary troilite grains within 15 ferromagnesian chondrules (10 FeO-poor and 5 FeO-rich chondrules) in the least metamorphosed ordinary chondrites, Bishunpur and Semarkona, have been measured by ion microprobe. Some troilite grains are located inside metal spherules within chondrules. Since such an occurrence is unlikely to be formed by secondary sulfidization processes in the solar nebula or on parent bodies, those troilites are most likely primary, having survived chondrule-forming high-temperature events. If they are primary, they may be the residues of evaporation at high temperatures during chondrule formation and may have recorded mass-dependent isotopic fractionations. However, the supposed primary troilites measured in this study do not show any significant sulfur isotopic fractionations (<1 ‰/amu) relative to large troilite grains in matrix. Among other chondrule troilites that we measured, only one (BI-CH22) apparently has a small excess of heavy isotopes (2.7 ± 1.4 ‰/amu) consistent with isotopic fractionation during evaporation. All other grains have isotopic fractionations of <1 ‰/amu. Because sulfur is so volatile that evaporation during chondrule formation is probably inevitable, non-Rayleigh evaporation most likely explains the lack of isotopic fractionation in putative primary troilite inside chondrules. Evaporation through the surrounding silicate melt would have suppressed the isotopic fractionation after silicate dust grains melted. At lower temperatures below extensive melting of silicates, a heating rate of >10⁴-10⁶ K/h would be required to avoid a large degree of sulfur isotopic fractionation in the chondrule precursors. This heating rate may provide a new constraint on the chondrule formation processes.     

Chondrule reheating experiments and relict olivine

Chondrules contain foreign objects, including some olivine grains that obviously did not crystallize from their silicate melt. The term recycling is usually applied to chondrules with relict grains, implying that the precursor contained relicts of a previous generation of chondrules. This has given rise to the idea that the pervasive melt droplet formation that affected the early solar system involved repeated events in which chondrules or chondrule debris were reheated. We conducted experiments in which synthetic chondrules generated from fine-grained mineral aggregates were heated and cooled a second time to see what the textural consequences of this reheating would be. Charges were heated to peak temperatures for 1 min and were cooled to near-solidus temperatures over 35 min, for both thermal cycles. We first made microporphyritic olivine charges and on reheating and second cooling observed coarser grain sizes and disappearance of relict grains, if the second peak temperature was the same as or higher than the first (but insufficient for destroying all nuclei). The coarsening was due to the dissolution of the smallest first generation crystals and additional growth on the relicts during cooling. Reheated barred olivine spheres generated barred olivine spheres again, no matter how low the peak temperature. This is because the number of remaining olivine grains or nuclei that acted as sites for regrowth was constant. Generating the observed distribution of chondrule textures, dominantly porphyritic, directly from a fine-grained precursor such as nebular or presolar condensates is impossible with a single event. With reheating of chondrules, generating the texture distribution is possible provided that subsequent heating events have higher peak temperatures than the first, so that total dissolution of the smallest grains occurs, with consequent coarsening. For our thermal history and a reasonable distribution of peak temperatures, multiple recycling events might be needed to make most chondrules porphyritic. Alternatively, the predominance of porphyritic textures in chondrules could be explained by heating times hours long for a fine-grained precursor or by heating of a coarse-grained precursor.The presence of relict grains derived from older chondrules or other material suggests that an aggregate has been heated for the first time, because recycling brings an approach to equilibrium. There appears to be no reliable way to use textures to tell just how many chondrules have been heated more than once. The relict grains simply indicate the nature of the precursors, which were at least in part derived from earlier chondrules, and of the peak temperatures too low for total melting and heating times too short for total dissolution. Rim thicknesses on relict grains depend on number density of crystals and melt composition, and are not a reliable guide to the chondrule cooling rate.     

Petrogenesis of Al-rich chondrules: Evidence from bulk compositions and phase equilibria

We measured major, minor, and trace-element compositions for eleven Al-rich chondrules from unequilibrated ordinary chondrites to investigate the relationships between Al-rich chondrules, ferromagnesian chondrules, Ca-, Al-rich inclusions (CAIs), and amoeboid olivine aggregates (AOAs). Phase equilibrium considerations show that, for the most part, mineral assemblages in Al-rich chondrules are those expected from melts of the observed compositions. The diversity of mineral assemblages and Al-rich chondrule types arises mainly from the fact that the array of compositions spans both the spinel-saturated anorthite-forsterite reaction curve and a thermal divide defined by where the anorthite-forsterite join crosses the reaction curve. The reaction curve accounts for the two principal varieties of Al-rich chondrule, plagioclase-phyric and olivine-phyric, with or without aluminous spinel. The thermal divide influences the subsequent evolution of each variety. A third variety of Al-rich chondrule contains abundant sodium-rich glass; trace-element fractionation patterns suggest that these glassy Al-rich chondrules could have been derived from the other two by extensive alteration of plagioclase to nepheline followed by remelting. The bulk compositions of Al-rich chondrules (except sodium-rich ones) are intermediate in a volatility sense between ferromagnesian chondrules and type C CAIs. The combined trend of bulk compositions for CAIs, Al-rich chondrules, and ferromagnesian chondrules mirrors, but does not exactly match, the trend predicted from equilibrium condensation at P_T ∼ 10^-3 atm; the observed trend does not match the trend found for evaporation from a liquid of chondritic composition. We thus infer that the bulk compositions of the precursors to CAIs, Al-rich chondrules, were ferromagnesian chondrules were controlled primarily by vapor-solid reactions (condensation or sublimation) in the solar nebula. Some Al-rich chondrules are consistent with an origin by melting of a compound CAI-ferromagnesian chondrule hybrid; others cannot be so explained. Any hybrid model is restricted by the constraint that the CAI precursor consisted dominantly of pyroxene + plagioclase + spinel; melilite cannot have been a significant component. Amoeboid olivine aggregates also have the inferred mineralogical characteristics of Al-rich chondrule precursors—they are mixtures of olivine with plagioclase-spinel-pyroxene-rich CAIs—but the few measured bulk compositions are more olivine-rich than those of Al-rich chondrules.     

Ca,Al-rich inclusions, amoeboid olivine aggregates, and Al-rich chondrules from the unique carbonaceous chondrite Acfer 094: I. mineralogy and petrology

Based on their mineralogy and petrography, ∼200 refractory inclusions studied in the unique carbonaceous chondrite, Acfer 094, can be divided into corundum-rich (0.5%), hibonite-rich (1.1%), grossite-rich (8.5%), compact and fluffy Type A (spinel-melilite-rich, 50.3%), pyroxene-anorthite-rich (7.4%), and Type C (pyroxene-anorthite-rich with igneous textures, 1.6%) Ca,Al-rich inclusions (CAIs), pyroxene-hibonite spherules (0.5%), and amoeboid olivine aggregates (AOAs, 30.2%). Melilite in some CAIs is replaced by spinel and Al-diopside and/or by anorthite, whereas spinel-pyroxene assemblages in CAIs and AOAs appear to be replaced by anorthite. Forsterite grains in several AOAs are replaced by low-Ca pyroxene. None of the CAIs or AOAs show evidence for Fe-alkali metasomatic or aqueous alteration. The mineralogy, textures, and bulk chemistry of most Acfer 094 refractory inclusions are consistent with their origin by gas-solid condensation and may reflect continuous interaction with SiO and Mg of the cooling nebula gas. It appears that only a few CAIs experienced subsequent melting. The Al-rich chondrules (ARCs; >10 wt% bulk Al₂O₃) consist of forsteritic olivine and low-Ca pyroxene phenocrysts, pigeonite, augite, anorthitic plagioclase, ± spinel, FeNi-metal, and crystalline mesostasis composed of plagioclase, augite and a silica phase. Most ARCs are spherical and mineralogically uniform, but some are irregular in shape and heterogeneous in mineralogy, with distinct ferromagnesian and aluminous domains. The ferromagnesian domains tend to form chondrule mantles, and are dominated by low-Ca pyroxene and forsteritic olivine, anorthitic mesostasis, and Fe,Ni-metal nodules. The aluminous domains are dominated by anorthite, high-Ca pyroxene and spinel, occasionally with inclusions of perovskite; have no or little FeNi-metal; and tend to form cores of the heterogeneous chondrules. The cores are enriched in bulk Ca and Al, and apparently formed from melting of CAI-like precursor material that did not mix completely with adjacent ferromagnesian melt. The inferred presence of CAI-like material among precursors for Al-rich chondrules is in apparent conflict with lack of evidence for melting of CAIs that occur outside chondrules, suggesting that these CAIs were largely absent from chondrule-forming region(s) at the time of chondrule formation. This may imply that there are several populations of CAIs in Acfer 094 and that mixing of “normal” CAIs that occur outside chondrules and chondrules that accreted into the Acfer 094 parent asteroid took place after chondrule formation. Alternatively, there may have been an overlap in the CAI- and chondrule-forming regions, where the least refractory CAIs were mixed with Fe-Mg chondrule precursors. This hypothesis is difficult to reconcile with the lack of evidence of melting of AOAs which represent aggregates of the least refractory CAIs and forsterite grains.     

The conditions of chondrule formation, Part I: Closed system

Mixing was an important process in the early solar nebula and is often used as an argument to explain the compositional scatter among chondrules—mm-sized, once molten silicate spherules in chondritic meteorites. If it is hypothesized that chondrules only acted as closed systems and the scatter in chondrule bulk chemical compositions is only the result of mixing heterogeneous precursor grains—the basic components of chondrules—, it is in turn possible to determine the sizes of the precursor grains using statistical calculations. In order to reproduce the observed compositional scatter in chondrules not more than ∼10 precursor grains should contribute to a single chondrule, each with a diameter of several 100 μm. This finding has important implications for the conditions of chondrule formation and replaces the so far widely accepted model that chondrules formed from fine-grained “dust-balls”. Chondrules rather formed from coarse-grained precursor aggregates with variable amounts of μm-fine matrix material. As a consequence, only chondrite matrix or interstellar material winnows as precursor material. Large grains of variable composition serving as precursor grains must have been formed prior to chondrule formation. Chondrules probably have not been their immediate precursors, as only 1-2 chondrule recycling steps would have homogenized bulk chondrule compositions. Chondrule recycling can therefore only have occurred to a limited extent. Chondrule formation needed at least three steps: (1) production of large and heterogeneous chondrule precursor grains, (2) agglomeration of large precursor grains and fine-grained precursors into aggregates, (3) formation of chondrules during transient heating events. Al-rich chondrules can in this context be explained by the admixture of CAIs to either chondrule precursors or a population of existing chondrules.     

Physical properties of chondrules in different chondrite groups: Implications for multiple melting events in dusty environments

Chondrite groups (CV, CK, CR) with large average chondrule sizes have low proportions of RP plus C chondrules, high proportions of enveloping compound chondrules, high proportions of chondrules with (thick) igneous rims, and relatively low proportions of type-I chondrules containing sulfide. In contrast, chondrite groups (CM, CO, OC, R, EH, EL) with smaller average chondrule sizes have the opposite properties. Equilibrated CK chondrites have plagioclase with relatively low Na; equilibrated OC, R, EH and EL chondrites have more sodic plagioclase. Enveloping compound chondrules and chondrules with igneous rims formed during a remelting event after the primary chondrule was incorporated into a dustball. Repeated episodes of remelting after chondrules were surrounded by dust would tend to produce large chondrules. RP and C chondrules formed by complete melting of their precursor assemblages; remelting of RP and C chondrules surrounded by dust would tend to produce porphyritic chondrules as small dust particles mixed with the melt, providing nuclei for crystallizing phenocrysts. This process would tend to diminish the numbers of RP and C chondrules. Correlations among these chondrule physical properties suggest that chondrite groups with large chondrules were typically surrounded by thick dust-rich mantles that formed in locally dusty nebular environments. Chondrules that were surrounded by thick dust mantles tended to cool more slowly because heat could not quickly radiate away. Slow cooling led to enhanced migration of sulfide to chondrule surfaces and more extensive sulfide evaporation. These chondrules also lost Na; the plagioclase that formed from equilibrated CK chondrites was thus depleted in Na.     

Oxygen isotope compositions of chondrules in CR chondrites

We report in situ ion microprobe analyses of oxygen isotopic compositions of olivine, low-Ca pyroxene, high-Ca pyroxene, anorthitic plagioclase, glassy mesostasis, and spinel in five aluminum-rich chondrules and nine ferromagnesian chondrules from the CR carbonaceous chondrites EET92042, GRA95229, and MAC87320. Ferromagnesian chondrules are isotopically homogeneous within ±2‰ in Δ¹⁷O; the interchondrule variations in Δ¹⁷O range from 0 to −5‰. Small oxygen isotopic heterogeneities found in two ferromagnesian chondrules are due to the presence of relict olivine grains. In contrast, two out of five aluminum-rich chondrules are isotopically heterogeneous with Δ¹⁷O values ranging from −6 to −15‰ and from −2 to −11‰, respectively. This isotopic heterogeneity is due to the presence of ¹⁶O-enriched spinel and anorthite (Δ¹⁷O = −10 to −15‰), which are relict phases of Ca,Al-rich inclusions (CAIs) incorporated into chondrule precursors and incompletely melted during chondrule formation. These observations and the high abundance of relict CAIs in the aluminum-rich chondrules suggest a close genetic relationship between these objects: aluminum-rich chondrules formed by melting of spinel-anorthite-pyroxene CAIs mixed with ferromagnesian precursors compositionally similar to magnesium-rich (Type I) chondrules. The aluminum-rich chondrules without relict CAIs have oxygen isotopic compositions (Δ¹⁷O = −2 to −8‰) similar to those of ferromagnesian chondrules. In contrast to the aluminum-rich chondrules from ordinary chondrites, those from CRs plot on a three-oxygen isotope diagram along the carbonaceous chondrite anhydrous mineral line and form a continuum with amoeboid olivine aggregates and CAIs from CRs. We conclude that oxygen isotope compositions of chondrules resulted from two processes: homogenization of isotopically heterogeneous materials during chondrule melting and oxygen isotopic exchange between chondrule melt and ¹⁶O-poor nebular gas.     

High precision SIMS oxygen three isotope study of chondrules in LL3 chondrites: Role of ambient gas during chondrule formation

We report high precision SIMS oxygen three isotope analyses of 36 chondrules from some of the least equilibrated LL3 chondrites, and find systematic variations in oxygen isotope ratios with chondrule types. FeO-poor (type I) chondrules generally plot along a mass dependent fractionation line (Δ¹⁷O ∼ 0.7‰), with δ¹⁸O values lower in olivine-rich (IA) than pyroxene-rich (IB) chondrules. Data from FeO-rich (type II) chondrules show a limited range of δ¹⁸O and δ¹⁷O values at δ¹⁸O = 4.5‰, δ¹⁷O = 2.9‰, and Δ¹⁷O = 0.5‰, which is slightly ¹⁶O-enriched relative to bulk LL chondrites (Δ¹⁷O ∼ 1.3‰). Data from four chondrules show ¹⁶O-rich oxygen isotope ratios that plot near the CCAM (Carbonaceous Chondrite Anhydrous Mineral) line. Glass analyses in selected chondrules are systematically higher than co-existing minerals in both δ¹⁸O and Δ¹⁷O values, whereas high-Ca pyroxene data in the same chondrule are similar to those in olivine and pyroxene phenocrysts.Our results suggest that the LL chondrite chondrule-forming region contained two kinds of solid precursors, (1) ¹⁶O-poor precursors with Δ¹⁷O > 1.6‰ and (2) ¹⁶O-rich solid precursors derived from the same oxygen isotope reservoir as carbonaceous chondrites. Oxygen isotopes exhibited open system behavior during chondrule formation, and the interaction between the solid and ambient gas might occur as described in the following model. Significant evaporation and recondensation of solid precursors caused a large mass-dependent fractionation due to either kinetic or equilibrium isotope exchange between gas and solid to form type IA chondrules with higher bulk Mg/Si ratios. Type II chondrules formed under elevated dust/gas ratios and with water ice in the precursors, in which the ambient H₂O gas homogenized chondrule melts by isotope exchange. Low temperature oxygen isotope exchange may have occurred between chondrule glasses and aqueous fluids with high Δ¹⁷O (∼5‰) in LL the parent body. According to our model, oxygen isotope ratios of chondrules were strongly influenced by the local solid precursors in the proto-planetary disk and the ambient gas during chondrule melting events.     

An Fe isotope study of ordinary chondrites

The Fe isotope composition of ordinary chondrites and their constituent chondrules, metal and sulphide grains have been systematically investigated. Bulk chondrites fall within a restricted isotopic range of <0.2‰ δ⁵⁶Fe, and chondrules define a larger range of >1‰ (−0.84‰ to 0.21‰ relative to the IRMM-14 Fe standard). Fe isotope compositions do not vary systematically with the very large differences in total Fe concentration, or oxidation state, of the H, L, and LL chondrite classes. Similarly, the Fe isotope compositions of chondrules do not appear to be determined by the H, L or LL classification of their host chondrite. This may support an origin of the three ordinary chondrite groups from variable accretion of identical Fe-bearing precursors.A close relationship between isotopic composition and redistribution of Fe during metamorphism on ordinary chondrite parent bodies was identified; the largest variations in chondrule compositions were found in chondrites of the lowest petrologic types. The clear link between element redistribution and isotopic composition has implications for many other non-traditional isotope systems (e.g. Mg, Si, Ca, Cr). Isotopic compositions of chondrules may also be determined by their melting history; porphyritic chondrules exhibit a wide range in isotope compositions whereas barred olivine and radial pyroxene chondrules are generally isotopically heavier than the ordinary chondrite mean. Very large chondrules preserve the greatest heterogeneity of Fe isotopes.The mean Fe isotope composition of bulk ordinary chondrites was found to be −0.06‰ (±0.12‰ 2 SD); this is isotopically lighter than the terrestrial mean composition and all other published non-chondritic meteorite suites e.g. lunar and Martian samples, eucrites, pallasites, and irons. Ordinary chondrites, though the most common meteorites found on Earth today, were not the sole building blocks of the terrestrial planets.     

Experimental study of evaporation and isotopic mass fractionation of potassium in silicate melts

The behavior of Na and K during evaporation from chondrule composition melts was studied using a vacuum furnace. Though Na is the less volatile of the two as an element, it is lost more rapidly than K from silicate melts. Mass fractionation of K isotopes was measured by ion microprobe and Rayleigh fractionation is observed for vacuum evaporation (10⁻⁵ atm). With higher pressures of air, the K loss rate decreases but with increasing hydrogen pressure, K is lost more rapidly. δ⁴¹K decreases with higher pressures, because of back reaction between melt and K in the gas. With long heating duration, the release of light K condensed within the furnace leads to interaction with the K-depleted melt and a further reduction of δ⁴¹K. Natural chondrules differ in some ways from our experimental residues. Some (especially type IIA) have superchondritic Na and K, despite their assumed formation in nebular hydrogen, which enhances volatile loss, and chondrules do not show K isotopic fractionation. Type I chondrules in Semarkona (LL3.0) either plot on our evaporation trend, or are depleted in K but slightly enriched in Na, relative to K. In Bishunpur (LL3.1), type I chondrules are mostly K-depleted but moderately to strongly enriched in Na. In petrologic type 3.2 to 3.4 chondrites they are enriched in both K and Na, like type II chondrules. The alkali contents suggest type I chondrules experienced evaporation and subsequent metasomatism. Their normal δ⁴¹K values suggest closed-system evaporation of a chondritic precursor in a gas with relatively high K pressures due to vaporization of dust accompanying chondrule precursor aggregates. Type II chondrules are volatile-rich, as well as unfractionated in K isotopes. They probably formed in a gas with higher pK than in the case of type I chondrules, due to heating of a more dust-rich parcel of gas.     

An experimental study of the formation of metallic iron in chondrules

The abundance of metallic iron is highly variable in different kinds of chondrites. The precise mechanism by which metal fractionation occurred and its place in time relative to chondrule formation are unknown. As metallic iron is abundant in most Type I (FeO-poor) chondrules, determining under what conditions metal could form in chondrules is of great interest. Assuming chondrules were formed from low temperature nebular condensate, we heated an anhydrous CI-like material at 1580°C in conditions similar to those of the canonical nebula (P_H2 = 1.3 × 10⁻⁵ atm). We reproduced many of the characteristics of Type IA and IIA chondrules but none of them contained any iron metal. In these experiments FeO was abundant in charges that were heated for as long as 6 h. At a lower temperature, 1350°C, dendritic/cellular metal crystallized from Fe-FeS melts during the evaporation of S. However, the silicate portion consisted of many relict grains and vesicles, not typical of chondrules.Evaporation experiments conducted at P_H2 = 1 atm and 1565°C produced charges containing metallic iron both as melt droplets and inclusions in olivine, similar to those found in chondrules. Formation of iron in these experiments was primarily the result of desulfurization of FeS. With long heating times Fe° was lost by evaporation. Apart from some reduction of FeO by kerogen to make metal inclusions within olivine grains, reduction of FeO to make Fe° in these charges was not observed.This study shows that under canonical nebular conditions FeS and iron-metal are extremely volatile so that metal-rich Type I chondrules could not form by melting “CI.” Under these conditions FeO is lost predominantly by hydrogen stripping and, due to the relative low abundance of hydrogen at low pressures, remains in the melt for as long as 6 h. Conversely, at higher total pressures (1-atm H₂) iron metal (produced mainly by the desulfurization of troilite) is less volatile and remains in the melt for longer times (at least 6 h). In addition, due to elevated pressures of hydrogen, FeO is stripped away much faster. These results suggest that chondrule formation occurred in environments with elevated pressures relative to the canonical nebula for iron metal to be present.     

Heating experiments relevant to the depletion of Na,K and Mn in the Earth and other planetary bodies

We have studied the evaporation of Na, K and Mn from Al-Na-K- and Mn-rich silicates at various conditions. Total alkali oxide contents ranged from 5 to 20%. The evaporation rate of Na increases with temperature and decreasing oxygen fugacity and decreases with duration of heating. The loss of K is in all cases less pronounced than for Na. Heating in an evacuated vacuum furnace is more effective in removing Na and K from melt droplets than in furnaces with one atm gas flow of air or gas mixtures controlling the oxygen fugacity. The strong pumping required to keep the vacuum removes Na and K atoms very effectively. In all experiments, the rate of evaporation is determined by quasi-equilibrium between a thin layer of Na and K rich gas above the molten silicates. The results of the experiments are in agreement with several other studies.In experiments with more than one sample in the furnace, equilibration of Na- and K-rich samples with Na- and K-poor samples occurred rapidly, mediated by the ambient gas phase.The results of experiments with Mn in starting compositions showed much stronger losses of Na than Mn under a variety of conditions.Thus the nearly chondritic Mn/Na ratios in the Earth cannot be the result of evaporation of Na and Mn in Earth-making materials, as the Mn/Na ratios in evaporation residues would be much higher than chondritic ratios. Such evaporation processes may have occurred in the parent material of Moon, Vesta and Mars.The data suggest, in agreement with earlier hypotheses, that the high and variable contents of Na and K in chondrules require a gas phase high in Na and K equilibrating with chondrule melts. The volume of nebular gas parental to a certain type of chondrites was heated and Na and K were lost from the chondrule precursors to the gas phase. Subsequently the nebular parcel was compressed leading to higher partial pressures of Na and K. Flash heating then produced chondrule melts which incorporated some of the gaseous Na and K and then cooled rapidly. The large range of Na and K contents in chondrule melts reflects very local enrichments of Na and K in the gas phase. Despite these variations bulk chondritic meteorites have well defined bulk Na and K contents, implying a closed system during formation of chondrules and matrix.     

Evaporation and recondensation of sodium in Semarkona Type II chondrules

We have investigated the Na distributions in Semarkona Type II chondrules by electron microprobe, analyzing olivine and melt inclusions in it, mesostasis and bulk chondrule, to see whether they indicate interactions with an ambient gas during chondrule formation. Sodium concentrations of bulk chondrule liquids, melt inclusions and mesostases can be explained to a first approximation by fractional crystallization of olivine ± pyroxene. The most primitive olivine cores in each chondrule are mostly between Fa₈ and Fa₁₃, with 0.0022–0.0069 ± 0.0013 wt.% Na₂O. Type IIA chondrule olivines have consistently higher Na contents than olivines in Type IIAB chondrules. We used the dependence of olivine–liquid Na partitioning on FeO in olivine as a measure of equilibration. Extreme olivine rim compositions are ～Fa₃₅ and 0.03 wt.% Na₂O and are close to being in equilibrium with the mesostasis glass. Olivine cores compared with the bulk chondrule compositions, particularly in IIA chondrules, show very high apparent D_Na, indicating disequilibrium and suggesting that chondrule initial melts were more Na-rich than present chondrule bulk compositions. The apparent D_Na values correlate with the Na concentrations of the olivine, but not with concentrations in the bulk melt. We use equilibrium D_Na to find the Na content of the true parent liquid and estimate that Type IIA chondrules lost more than half their Na and recondensation was incomplete, whereas Type IIAB chondrules recovered most of theirs in their mesostases.Glass inclusions in olivine have lower Na than expected from fractionation of bulk composition liquids, and mesostases have higher Na than expected in calculated daughter liquids formed by fractional crystallization alone. These observations also require open system behavior of chondrules, specifically evaporation of Na before formation of melt inclusions followed by recondensation of Na in mesostases. Within this record of evaporation followed by recondensation, there is no indication of a stage with zero Na in the chondrules, which is predicted by models for shock wave cooling at canonical nebular pressures, suggesting high P_T.The high Na concentrations in olivine and mesostases indicate very high P_Na while chondrules were molten. This may be explained by local, very high particle densities where Type II chondrules formed. The high P_T, P_Na and number densities of chondrules implied suggest formation in debris clouds after protoplanetary collisions as an alternative to formation after passage of shock waves through large particle-rich clumps in the disk. Encounters of partially molten chondrules should have been frequent in these dense swarms. However, in many ordinary chondrites like Semarkona, “cluster chondrites”, compound chondrules are not abundant but instead chondrules aggregated into clusters. Chondrule melting, cooling and clustering in dense swarms contributed to rapid accretion, possibly after collision, by fallback on the grandparent body and by reaccretion as a new body downrange.     

The origin of chondrules: experimental investigation of metastable liquids in the system Mg2SiO4-SiO2

Laser-melted magnesium silicate droplets, supercooled 400–750°C below their equilibrium liquidus temperatures before crystallization, were examined to provide a comparison with meteoritic and lunar chondrules and to examine physicochemical parameters that may indicate the conditions of their formation. Internal textures of the spherules strikingly resemble textures observed in some chondrules. Definite trends in crystal morphology, crystal width and texture were established with respect to nucleation temperature and bulk composition. Such trends provide a framework for determining the nucleation temperature of chondrules. The only phase to nucleate from the supercooled forsterite-enstatite normative melts was forsterite, which was present in more-than-normative amounts. Highly siliceous glass (~65wt. % SiO₂) was identified interstitially to the forsterite crystals in seven of the spherules and is thought to be present in all. The presence of enstatite and the large proportion of crystals in some meteoritic chondrules implies that they were maintained at temperatures considerably in excess of 600°C at some point in their history.     

Matrix material in type 3 chondrites—occurrence,heterogeneity and relationship with chondrules

Matrix material in type 3 chondrites forms rims on chondrules, metal-sulfide aggregates, Ca,Al-rich inclusions and chondritic clasts; it also forms lumps up to a millimeter in size, which may contain coarser silicates. Chondrules of all types were found with internal matrix lumps that appear to have entered the chondrules before the latter had crystallized. Mean concentrations of Mg, Na, Al and Ca in matrix occurrences show up to fivefold variations in a single chondrite. Variations between mean matrix compositions of individual type 3 ordinary chondrites are almost as large and partly reflect systematic differences between H, L and LL matrices. Such variations are probably a result of nebular separation of feldspathic material and ferromagnesian silicates.Compositions of chondrules and their matrix rims are normally unrelated, although rim compositions are correlated with those of matrix lumps inside chondrules. A single chondrule was found with a composition nearly identical to that of its internal matrix lump, suggesting that some chondrules may have formed from matrix material. Matrix lumps are as heterogeneous as chondrules, but mean chondrule and matrix compositions differ, even allowing for possible loss of metallic Fe,Ni during chondrule formation. Since bulk compositions of matrix lumps and rims have probably not changed significantly since their formation except for Fe-Mg exchange, our matrix samples cannot represent typical chondrule precursor materials.     

The petrology of chondrules in the sharps meteorite

Correlated petrographic and microprobe studies of 96 chondrules in the Sharps (H-3) chondrite indicate that chondritic material had a highly varied pre-accumulation history. Some chondrules, chiefly excentroradial and barred types, appear to be quenched droplets. Others, including most of the metal poor microporphyritic type, appear to have crystallized more slowly and are thought to be fragments of pre-existing rock. Although chondrules of all types show various effects similar to those produced by shock, such effects are most conspicuous in metal-rich chondrules and least conspicuous in spherical chondrules. It is concluded that shock was involved in the origin of chondrules and not simply a secondary effect.It is proposed that chondrules were formed by shock processes during the accumulation of nebular dust into asteroid-sized bodies. Olivine-rich microporphyritic chondrules are thought to be due to complete melting of large masses of target material; metal-rich chondrules represent shock melting and partial vaporization; and spherical, pyroxene-rich chondrules are interpreted as condensates from shock-generated vapor.     

Ubiquitous low-FeO relict grains in type II chondrules and limited overgrowths on phenocrysts following the final melting event

Type II porphyritic chondrules commonly contain several large (>40 μm) olivine phenocrysts; furnace-based cooling rates based on the assumption that these phenocrysts grew in a single-stage melting-cooling event yield chondrule cooling-rate estimates of 0.01-1 K s⁻¹. Because other evidence indicates much higher cooling rates, we examined type II chondrules in the CO3.0 chondrites that have experienced only minimal parent-body alteration. We discovered three kinds of evidence indicating that only minor (4-10 μm) olivine growth occurred after the final melting event: (1) Nearly all (>90%) type II chondrules in CO3.0 chondrites contain low-FeO relict grains; overgrowths on these relicts are narrow, in the range of 2-12 μm. (2) Most type II chondrules contain some FeO-rich olivine grains with decurved surfaces and acute angles between faces indicating that the grains are fragments from an earlier generation of chondrules; the limited overgrowth thicknesses following the last melting event are too thin to disguise the shard-like nature of these grains. (3) Most type II chondrules contain many small (<20 μm) euhedral or subhedral phenocrysts with central compositions that are much more ferroan than the centers of the large phenocrysts; their small sizes document the small amount of growth that occurred after the final melting event. If overgrowth thicknesses were small (4-10 μm) after the final melting event, it follows that large fractions of coarse (>40 μm) high-FeO phenocrysts are relicts from earlier generations of chondrules, and that cooling rates after the last melting event were much more rapid than indicated by models based on a single melting event. These observations are thus inconsistent with the “classic” igneous model of formation of type II porphyritic chondrules by near-total melting of a precursor mix followed by olivine nucleation on a very limited number of nuclei (say, ≤10) and by growth to produce the large phenocrysts during a period of monotonic (and roughly linear) cooling. Our observations that recycled chondrule materials constitute a large component of the phenocrysts of type II chondrules also imply that this kind of chondrule formed relatively late during the chondrule-forming period.     

Melting of carbon at 50 to 300 kbar

Diamond (～1 μm) and graphite (1–10 μm) in NaCl were melted at 50 to 300 kbar in a diamond anvil cell using a pulsed YAG laser. The samples were removed from the cell and the structures of the quenched phases were studied by transmission electron microscopy. The melted regions of the samples were found to consist of nearly perfect spheres of carbon ranging in size from ～1 μm down to less than a few nanometers. In the diamond sample melted at 300 kbar, the larger spherules (>0.2 μm) are polycrystalline diamond with either a granular or radial texture. The smaller spherules (<0.2 μm) give electron diffraction patterns with four diffuse rings that correspond to the 002, 100, 004 and 110 of graphite. This diffraction pattern is typical of disordered graphite randomly oriented about the c-axis. Dark field imaging, using a portion of the 002 ring, produces a “bow tie” figure in each of the smaller spherules. The orientation of the “bow tie” figure depends on the portion of the ring used to form the image, and indicates a radial orientation of the c-axis of the disordered graphite. The spacing between the 002 layers depends on the pressure at the time of melting. We interpret this to indicate that there is some sp³ bonding between layers in the disordered graphite in the smaller spherules. The smaller spherules may have the disordered graphite structure because of the effect of the size on the free energy relationship between the phases, or they may have been quenched more rapidly than the larger spherules thus preserving some of the character of the melt. If the latter explanation is correct, then our results may indicate that the diamond melt contains significant sp² bonding. Lattice images (Fig. 12) of the internal structure of the smallest spherules observable (～50 A) clearly show that the carbon layers are parallel to the surface of the spherules and that there is a great deal of disorder in the layers. These observations are entirely consistent with our conclusions based on the dark field images.