Laura L. Lundberg

Correlated study of initial 87Sr86Sr and AlMg isotopic systematics and petrologic properties in a suite of refractory inclusions from the Allende meteorite

We have studied a suite of six coarse-grained Ca-Al-rich inclusions (CAIs) from the Allende meteorite, applying in concert detailed petrographic and chemical characterization, ion probe mass spectrometric analyses of the Al-Mg isotopic system to determine the abundance and distribution of 26A1, and thermal emission mass spectrometric analyses of the Rb-Sr system to determine initial 87Sr86Sr ratios. The simplest inclusion is a Type Bl, USNM 3529-Z, that shows evidence of minor alteration and recrystallization since solidification from a melt droplet; e.g., some of the coarse-grained anorthite in the inclusion (especially near the inclusion rim) apparently formed by recrystallization or melting of fine-grained secondary anorthite that had replaced melilite. Four other Type Bs contain evidence of more widespread recrystallization in addition to evidence of local replacement of melilite by coarse anorthite; e.g., melilite zoning is complex and not igneous in origin, and in one inclusion (USNM 3529-21) melilite and pyroxene crystals enclose relicts of an earlier generation of the same phases. The sixth inclusion, USNM 3898, is a Type A whose properties appear to be largely metamorphic in origin. All six inclusions show variations in Mg isotopic compositions indicating that they incorporated 26A1. In 3529-Z the relative abundance of 26A1 (26Al27Al = (4.0 ± 0.1) × 10−5) is close to the “canonical” value (26Al27Al = 5 × 10−5) for CAIs, but both anorthite and melilite show minor departures from a strict isochron relationship between Mg and Al. In particular, the “coarse” anorthite near the inclusion rim has initial 26Al27Al ratios of 1−2 × 10−5, implying formation at least 1.5 million years after the “first-generation” anorthite. The other four Type Bs exhibit larger disturbances of the Al-Mg system that can be understood in terms of local recrystallization and isotopic exchange, primarily between anorthite and melilite. Only 3898 is consistent with a strict Al-Mg isochron relationship, but this CAI lacks primary anorthite, and the small range of Al/Mg among the phases analyzed may preclude resolution of any small isotopic disturbances. The RbSr isotopic system also typically indicates some level of disturbance, some of which must have been relatively recent. Nevertheless, phases characterized by very low Rb/Sr permit precise identification of initial 87Sr86Sr. Our Al-Mg and Sr data are largely consistent with a simple chronological interpretation of both systems; i.e., the 26Al results suggest that all the initial 87Sr86Sr ratios should have the same value and, with one exception, our initial 87Sr86Sr ratios are consistent with a single value; this value is somewhat higher than the primitive value ALL reported by Gray et al. (1973), however. Also with one exception, the very low Rb/Sr CAI data reported by Gray et al. (1973) are consistent with the same value. The exception is the same inclusion in both studies; our analysis of USNM 3898 yields an initial 87Sr86Srhigher than that of the other CAIs, while Gray et al. (1973) obtained an initial 87Sr86Srlower than that of the other CAIs for the same inclusion (labeled as D7, on which ALL is based). Neither result can be readily explained as a chronological aberration, since 3898 has nearly canonical 26Al27Al. The present data are the first which establish a firm association between primitive 26Al27Al and primitive 87Sr86Sr by analysis of both isotopic systems in the same inclusions. The comparison of both isotopic systems, interpreted as simple nebular chronometers, does not reveal any chronological inconsistencies that demand resolution in terms of a grossly heterogeneous distribution of 26Al.

Rare earth elements in minerals of the ALHA77005 shergottite and implications for its parent magma and crystallization history

Abstract Rare earth element (REE) and selected trace element abundances were measured in individual grains of pyroxenes, whitlockite, maskelynite and olivine of the ALHA77005 shergottite. Minor (~ 1%) whitlockite, which displays a negative Eu anomaly indicative of co-crystallization with plagioclase, dominates the bulk-rock REE pattern. Detectable Ce anomalies in ~20% of the pyroxene grains analyzed appear to be the result of weathering in the Antarctic environment and indicate that pyroxene is particularly susceptible to element remobilization. Mineral compositions and textural observations suggest that ALHA77005 is a cumulate with ~50% cumulus material (olivine + chromite). Olivine re-equilibrated with intercumulus liquid on cooling. The intercumulus liquid then crystallized as a closed system to form successively poikilitic low-Ca and high-Ca pyroxene, followed by interstitial pyroxenes, plagioclase, whitlockite (and other minor phases), and reacted with cumulus crystals to form rims of different composition. This crystallization sequence provides a constraint on the shergottite source region composition. The ALHA77005 and Shergotty intercumulus liquids had similar REE compositions; however, as opposed to Shergotty, if ALHA77005 was formed during the last 0.72 Ga, its parent melt was not contaminated by LREE-enriched material.

Rare earth element carriers in the Shergotty meteorite and implications for its chronology

Ion probe measurements of the rare earth element (REE) concentrations of individual grains of the Shergotty meteorite are reported. Phases analyzed include whitlockite, apatite, baddeleyite, augite, pigeonite, maskelynite and K-rich glass. U concentrations of whitlockite and apatite crystals were also obtained. The whole rock REE pattern is dominated by whitlockite, which contains over 95% of the light rare earth elements (LREE). REE concentrations in apatite are much lower than estimated by Laulet al. (1986). All the whitlockites (whether intergrown with pyroxene, equant or interstitial) have the same relative abundances of LREE (i.e. patterns are almost flat from La to Sm). The observation, by Joneset al. (1985), of a skeletal whitlockite with LREE enrichment is not confirmed by analyses of the same grain. Pyroxene rims are not enriched in LREE. No leachable carrier, enriched in LREE and associated with pyroxene (Laulet al., 1986; Jagoutz and Wanke, 1986), has been found. Instead, either a laboratory contamination or a petrographically cryptic “phase” such as a film on grain boundaries is suspected as the carrier of LREE enrichments. If a grain boundary film carries the enrichments, it would not be resistant to the metamorphic resetting which has affected other isotopic reservoirs in this sample. Thus, there is no compelling reason to consider the Sm-Nd pyroxene/leachates line (Jagoutz and Wanke, 1986) as a 360 m.y. isochron. Estimates of REE abundances in the Shergotty intercumulus melt indicate that a complex petrogenesis is required, in agreement with the conclusions of McKayet al. (1986a). Pyroxene distribution coefficients measured experimentally (McKayet al., 1986a) are compared with estimates from measured REE abundances in augite and pigeonite. Evolution of REE abundances in the Shergotty late-stage interstitial melt, as inferred from analyses of whitlockite, conforms with trends predicted from partitioning considerations, and requires no special processes such as metasomatism. The average U concentrations of whitlockite and apatite are respectively 540 and 1,550 ppb. Although the calcium phosphates are enriched in U, they contain less than 20% of the U in Shergotty.

Plutonium, uranium and rare earths in the phosphates of ordinary chondrites—the quest for a chronometer

Abstract The distributions of Pu, U and the REE (in particular Nd) in single crystals of the calcium phosphates, merrillite and apatite, of ordinary chondrites were studied to establish whether 244 Pu can be used to determine time differences in meteorite formation. Uranium was measured by an induced-fission track technique and the REE by secondary ion mass spectrometry. Limits on Pu concentrations (at the time of track retention) were inferred from fossil track measurements. It is argued here that most previous estimates of Pu concentrations, obtained from fossil tracks in phosphates, failed to correct adequately for spallation-induced tracks and were probably not correct. Plutonium is preferentially enriched in merrillite relative to apatite. In contrast, U concentrations are on the average 15 times higher in apatite than in merrillite. There are grain to grain variations, in the same phosphate phase from a given meteorite, for both elements. On the other hand, in the merrillite of a given meteorite, the REE concentrations are remarkably constant (typical grain to grain variations for type 5 and 6 chondrites are in most cases less than 10%). Average Nd concentrations in the merrillite of 15 different chondrites range from 96 to 164 ppm. REE abundances in apatite show more variations. Average Nd concentrations in the apatite of 13 chondrites range from 8 to 54 ppm and the ratios of Nd concentrations in merrillite relative to apatite range from 2.7 to 11. Despite the higher affinity of both Pu and the REE for merrillite than for apatite, there is no quantitative correlation between the abundances of these elements in merrillite grains from a given ordinary chondrite. Because of this lack of geochemical coherence between Pu and any of the REE, it is concluded that 244 Pu cannot be used to determine the relative formation times of chondrites. Bulk measurements that avoid the problems associated with fossil track measurements are discussed, however they have so far proven to be of little utility.

The evolution of a complex type B allende inclusion: An ion microprobe trace element study

Abstract USNM 5241 is a Type B1 refractory inclusion from Allende, first described by El Goresy et al. (1985), that consists of a 1.2 mm-thick melilite-rich and spinel-poor mantle enclosing a 0.6 cm-radius spinel-rich core; the inclusion contains xenoliths of spinel-free fassaite ± melilite ± anorthite incorporated within the spinel-rich core. Detailed ion microprobe analyses of individual phases in 5241 show that the rare earth element (REE) concentrations in mantle melilite vary irregularly with increasing distance from the rim of the inclusion, at first decreasing immediately below the rim and then remaining constant between ~0.4 and 1.0 mm. More than 1.0 mm from the rim, the REE concentrations again decrease. Although counterintuitive in the context of traditional fractional crystallization models, these REE variations are in fact broadly consistent with such a model in light of recent experimental measurements of DREE3+ (mel) by BECKETT et al. (1988) that show a strong inverse correlation of D with the akermanite content of the melilite. Local variations, over distances of We interpret 5241 as having formed largely by fractional crystallization during the first ~40% of its solidification; this was followed by fractional crystallization + xenolith assimilation during the last 60%.