Late-stage Accretion Research Articles

The non-carbonaceous nature of Earth’s late-stage accretion

Constraining the origin of Earth’s building blocks requires knowledge of the chemical and isotopic characteristics of the source region(s) where these materials accreted. The siderophile elements Mo and Ru are well suited to investigating the mass-independent nucleosynthetic (i.e., “genetic”) signatures of material that contributed to the latter stages of Earth’s formation. Studies contrasting the Mo and Ru isotopic compositions of the bulk silicate Earth (BSE) to genetic signatures of meteorites, however, have reported conflicting estimates of the proportions of the non-carbonaceous type or NC (presumptive inner Solar System origin) and carbonaceous chondrite type or CC (presumptive outer Solar System origin) materials delivered to Earth during late-stage accretion (likely including the Moon-forming event and onwards). The present study reports new mass-independent isotopic data for Mo, which are presumed to reflect the composition of the BSE. A comparison of the new estimate for the BSE composition with new data for a select suite of NC iron meteorites is used to constrain the genetic characteristics of materials accreted to Earth during late-stage accretion. Results indicate that the final 10 to 20 wt% of Earth’s accretion was dominated by NC materials that were likely sourced from the inner Solar System, although the addition of minor proportions of CC materials, as has been suggested to occur during accretion of the final 0.5 to 1 wt% of Earth’s mass, remains possible. If this interpretation is correct, it brings estimates of the genetic signatures of Mo and Ru during the final 10 to 20 wt% of Earth accretion into concordance.

Barrovian metamorphism in nominally lowermost Alpujarride Complex units: tectonic implications for the development of the orogenic wedge in the Western Mediterranean

ABSTRACT The sedimentary successions of the Permian-Triassic rifting related to the Alpine Tethys in the easternmost Lower Alpujarride units (Internal Betic Zone, south Spain) host shallow-intrusive basic bodies of Mesozoic age. These gabbroic bodies experienced low-grade Alpine metamorphic overprint, singularly recorded in three outcrops nearby Alicante city in southeast Spain: Callosa and Orihuela Ranges and Nueva Tabarca Island. Overall, the metabasic bodies still preserve magmatic ophitic/subophitic texture and relicts of magmatic augite and ferropargasite-ferroedenite-edenite and kaersutite. In contrast, metamorphic minerals include magnesioriebeckite, actinolite, albite, stilpnomelane, phengite, and chlorite. Major and trace element geochemical data suggest crystallization of the igneous protoliths from subalkaline (tholeiitic) to alkaline within-plate (continental) basaltic magmas slightly enriched in incompatible elements. These magmas with E-MORB geochemical affinity very likely intruded at shallow crustal levels in an extensional-synrift-geodynamic setting related to the breakoff of Pangea in the region. Pressure-Temperature (P-T) conditions estimated by average P-T method and pseudosection calculations suggest metamorphic peak at ca. 6 kbar at 340°C consistent with burial metamorphism down to ca. 18 km depth with intermediate thermal gradients of ca. 18°C/km. These conditions are different to the typical subduction-related blueschist facies conditions of the Lower Alpujarride units and suggest, instead, a collisional (Barrovian) event. This can be explained by a late-stage accretion of the studied units after the main subduction-related event, pointing to the dispersion of the corresponding Lower Alpujarride sections along the margin in different palaeogeographic positions before deformation and movement to their actual structural position. The metamorphic correlation of the studied units with the accreted passive margin of South Iberia (Subbetic zone) is also discussed.

Collision Chains among the Terrestrial Planets. II. An Asymmetry between Earth and Venus

Abstract During the late stage of terrestrial planet formation, hit-and-run collisions are about as common as accretionary mergers, for expected velocities and angles of giant impacts. Average hit-and-runs leave two major remnants plus debris: the target and impactor, somewhat modified through erosion, escaping at lower relative velocity. Here we continue our study of the dynamical effects of such collisions. We compare the dynamical fates of intact runners that start from hit-and-runs with proto-Venus at 0.7 au and proto-Earth at 1.0 au. We follow the orbital evolutions of the runners, including the other terrestrial planets, Jupiter, and Saturn, in an N-body code. We find that the accretion of these runners can take ≳10 Myr (depending on the egress velocity of the first collision) and can involve successive collisions with the original target planet or with other planets. We treat successive collisions that the runner experiences using surrogate models from machine learning, as in previous work, and evolve subsequent hit-and-runs in a similar fashion. We identify asymmetries in the capture, loss, and interchange of runners in the growth of Venus and Earth. Hit-and-run is a more probable outcome at proto-Venus, being smaller and faster orbiting than proto-Earth. But Venus acts as a sink, eventually accreting most of its runners, assuming typical events, whereas proto-Earth loses about half, many of those continuing to Venus. This leads to a disparity in the style of late-stage accretion that could have led to significant differences in geology, composition, and satellite formation at Earth and Venus.

On the formation of terrestrial planets between two massive planets: the case of 55 Cancri

ABSTRACT Considering the huge computational resources required by smoothed particle hydrodynamics (SPH) simulations and the overestimation of post-collision materials from perfect merging, we develop a statistical method to deal with collisions during the formation of planetary systems by introducing random material loss. In this method, the mass and water content lost by the sole outcome from every merger vary randomly within a range dependent on the total mass and water content of colliding bodies. The application of the random loss method to the planet formation in the Solar system shows a good consistency with existing SPH results. We also apply this method to the extrasolar planetary system 55 Cancri that hosts (at least) five planets and study the formation of terrestrial planets between the outermost two planets. A disc with 500 Mars mass embryos in dynamically cold orbits before the late-stage accretion phase is assumed. Scenarios with different amounts of planetary embryos and different loss parameters are adopted in our simulations. The statistical result from hundreds of simulations shows that an Earth-like planet with water inventory of roughly 6 Earth ocean could form between 55 Cnc f and d. It may reside between 1.0 and 2.6 au but the most likely region extends from 1.5 to 2.1 au. Thus, the probability of this planet being in the potentially habitable zone (0.59–1.43 au) is relatively low, only around 10 per cent. Planets 55 Cnc f and d could also be shaped and gain some water from giant impacts and consequently the orbits of them may also change accordingly.

Deciphering the origin of ionized gas in IC 1459 with VLT/MUSE

ABSTRACT IC 1459 is an early-type galaxy (ETG) with a rapidly counter-rotating stellar core, and is the central galaxy in a gas-rich group of spirals. In this work, we investigate the abundant ionized gas in IC 1459 and present new stellar orbital models to connect its complex array of observed properties and build a more complete picture of its evolution. Using the Multi-Unit Spectroscopic Explorer (MUSE), the optical integral field unit (IFU) on the Very Large Telescope (VLT), we examine the gas and stellar properties of IC 1459 to decipher the origin and powering mechanism of the galaxy’s ionized gas. We detect ionized gas in a non-disc-like structure rotating in the opposite sense to the central stars. Using emission-line flux ratios and velocity dispersion from full-spectral fitting, we find two kinematically distinct regions of shocked emission-line gas in IC 1459, which we distinguished using narrow (σ ≤ 155 km s−1) and broad (σ > 155 km s−1) profiles. Our results imply that the emission-line gas in IC 1459 has a different origin than that of its counter-rotating stellar component. We propose that the ionized gas is from late-stage accretion of gas from the group environment, which occurred long after the formation of the central stellar component. We find that shock heating and AGN activity are both ionizing mechanisms in IC 1459 but that the dominant excitation mechanism is by post-asymptotic giant branch stars from its old stellar population.

Nickel isotopic evidence for late-stage accretion of Mercury-like differentiated planetary embryos

Earth’s habitability is closely tied to its late-stage accretion, during which impactors delivered the majority of life-essential volatiles. However, the nature of these final building blocks remains poorly constrained. Nickel (Ni) can be a useful tracer in characterizing this accretion as most Ni in the bulk silicate Earth (BSE) comes from the late-stage impactors. Here, we apply Ni stable isotope analysis to a large number of meteorites and terrestrial rocks, and find that the BSE has a lighter Ni isotopic composition compared to chondrites. Using first-principles calculations based on density functional theory, we show that core-mantle differentiation cannot produce the observed light Ni isotopic composition of the BSE. Rather, the sub-chondritic Ni isotopic signature was established during Earth’s late-stage accretion, probably through the Moon-forming giant impact. We propose that a highly reduced sulfide-rich, Mercury-like body, whose mantle is characterized by light Ni isotopic composition, collided with and merged into the proto-Earth during the Moon-forming giant impact, producing the sub-chondritic Ni isotopic signature of the BSE, while delivering sulfur and probably other volatiles to the Earth.

Realistic collisional water transport during terrestrial planet formation

Context. According to the latest theoretical and isotopic evidence, Earth’s water content originates mainly from today’s asteroid belt region, or at least from the same precursor material. This suggests that water was transported inwards to Earth, and to similar planets in their habitable zone, via (giant) collisions of planetary embryos and planetesimals during the chaotic final phase of planet formation. Aims. In current dynamical simulations water delivery to terrestrial planets is still studied almost exclusively by assuming oversimplified perfect merging, even though water and other volatiles are particularly prone to collisional transfer and loss. To close this gap we have developed a computational framework to model collisional water transport by direct combination of long-term N-body computations with dedicated 3D smooth particle hydrodynamics (SPH) collision simulations of differentiated, self-gravitating bodies for each event. Methods. Post-collision water inventories are traced self-consistently in the further dynamical evolution, in accretionary or erosive as well as hit-and-run encounters with two large surviving bodies, where besides collisional losses, water transfer between the encountering bodies has to be considered. This hybrid approach enables us for the first time to trace the full dynamical and collisional evolution of a system of approximately 200 bodies throughout the whole late-stage accretion phase (several hundred Myr). As a first application we choose a Solar System-like architecture with already formed giant planets on either circular or eccentric orbits and a debris disk spanning the whole terrestrial planet region (0.5–4 au). Results. Including realistic collision treatment leads to considerably different results than simple perfect merging, with lower mass planets and water inventories reduced regularly by a factor of two or more. Due to a combination of collisional losses and a considerably lengthened accretion phase, final water content, especially with giant planets on circular orbits, is strongly reduced to more Earth-like values, and closer to results with eccentric giant planets. Water delivery to potentially habitable planets is dominated by very few decisive collisions, mostly with embryo-sized or larger objects and only rarely with smaller bodies, at least if embryos have formed throughout the whole disk initially. The high frequency of hit-and-run collisions and the differences to predominantly accretionary encounters, such as generally low water (and mass) transfer efficiencies, are a crucial part of water delivery, and of system-wide evolution in general.

Constraining Planetary Gas Accretion Rate from Hα Line Width and Intensity: Case of PDS 70 b and c

Abstract Recent observations of protoplanets embedded in circumstellar disks have shed light on the planet formation process. In particular, detection of hydrogen Balmer-line (Hα) emission gives direct constraints on late-stage accretion onto gas giants. Very recently Haffert et al. measured the spectral line widths, in addition to intensities, of emission from the two protoplanets orbiting PDS 70. Here, we study these protoplanets by applying radiation-hydrodynamic models of the shock-heated accretion flow onto protoplanets that Aoyama et al. have recently developed. As a result, we demonstrate that Hα line widths combined with intensities lead to narrowing down the possible ranges of the protoplanetary accretion rate and/or mass significantly. While the current spectral resolution is not high enough to derive a definite conclusion regarding their accretion process, high-resolution spectral imaging of growing protoplanets is highly promising.

Constraints on terrestrial planet formation timescales and equilibration processes in the Grand Tack scenario from Hf-W isotopic evolution

We examine 141 N-body simulations of terrestrial planet late-stage accretion that use the Grand Tack scenario, coupling the collisional results with a hafnium-tungsten (Hf-W) isotopic evolution model. Accretion in the Grand Tack scenario results in faster planet formation than classical accretion models because of higher planetesimal surface density induced by a migrating Jupiter. Planetary embryos that grow rapidly experience radiogenic ingrowth of mantle 182W that is inconsistent with the measured terrestrial composition, unless much of the tungsten is removed by an impactor core that mixes thoroughly with the target mantle. For physically Earth-like surviving planets, we find that the fraction of equilibrating impactor core kcore≥ 0.6 is required to produce results agreeing with observed terrestrial tungsten anomalies (assuming equilibration with relatively large volumes of target mantle material; smaller equilibrating mantle volumes would require even larger kcore). This requirement of substantial core re-equilibration may be difficult to reconcile with fluid dynamical predictions and hydrocode simulations of mixing during large impacts, and hence this result does not favor the rapid planet building that results from Grand Tack accretion.

Bulk chemical and Hf–W isotopic consequences of incomplete accretion during planet formation

Late-stage accretion involves collisions which may result in complete or incomplete merging of the two objects, hit-and-run encounters, or mass loss from the target. We use a recent N-body study incorporating these different collision styles (Chambers, J.E. [2013]. Icarus 224, 43–56) to investigate how collision style affects the bulk chemical and isotopic outcomes of terrestrial planet formation. Compared with simulations in which all collisions result in perfect mergers, the variability in modeled silicate mass fraction and tungsten isotope anomaly is larger, especially for lower-mass planets. The final tungsten anomaly also shows a systematic reduction, because the timescale to finish planet growth is longer when incomplete mergers are included. Simulations including incomplete merging can reproduce the observed scatter in both tungsten anomaly and silicate mass fraction of the terrestrial planets.

Partitioning of carbon between Fe-rich alloy melt and silicate melt in a magma ocean – Implications for the abundance and origin of volatiles in Earth, Mars, and the Moon

The budget and origin of carbon in Earth and other terrestrial planets are debated and one of the key unknowns is the fate of carbon during early planetary processes including accretion, core formation, and magma ocean (MO) crystallization. Here we determine, experimentally, the solubility of carbon in coexisting Fe–Ni alloy melt and basaltic silicate melt in shallow MO conditions, i.e., at 1–3GPa, 1500–1800°C. Oxygen fugacity of the experiments, estimated based on Fe (in metallic alloy melt)–FeO (in silicate melt) equilibrium, varied between ∼IW-0.4 and IW-1.0, where IW refers to the oxygen fugacity imposed by the coexistence of iron and wüstite. Four different starting mixes, each with 7:3 silicate:metal mass ratio and silicate melt NBO/T (estimated proportion of non-bridging oxygen with respect to tetrahedral cations; NBO/T=2×total OT-4, where T=Si+Ti+Al+Cr+P) ranging from 0.81 to 1.54 were studied. Concentrations of carbon in the alloy melt were determined using electron microprobe whereas carbon contents of quenched basaltic glasses were determined using secondary ionization mass spectrometry (SIMS). Identification of carbon and hydrogen-bearing species in silicate glasses was performed using Raman and Fourier Transformed Infrared (FTIR) spectroscopy.Our results show that carbon in the metallic melt varies between 4.4wt.% and 7.4wt.% and increases with increasing temperature and modestly with increasing pressure but decreases with increasing Ni content of the alloy melt. Carbon concentration in the silicate melts, on the other hand, varies from 11±1ppm to 111±7ppm and is negatively correlated with pressure but positively correlated with temperature, the NBO/T, the oxygen fugacity and the water content of the silicate melts. Raman and FTIR results show that at our experimental conditions, carbon in silicate melt is dissolved both as hydrogenated species and CO32-. The calculated carbon partition coefficient DCmetal/silicate varies from 510±53 to 5369±217 and varies systematically as a function of P, T, fO2, water content, the composition of the silicate melt (expressed using NBO/T), and Ni content of alloy melt (XNi). The range of DCmetal/silicate measured in our study with carbonated and hydrogenated carbon species in silicate melt is similar to that reported in the literature for experiments where carbonyl complexes are the chief carbon species in silicate melts. A parameterization was derived using the data from this and existing studies such as lnDCmetal/silicate=a/T+b·P/T+c·ln(fO2)+d·(NBO/T)+e·ln(1-XNi)+fwhere a=−33,510, b=1357, c=−0.596, d=−1.182, e=4.15, f=13.38, the temperature is in Kelvin, and the pressure is in gigapascal. Using this parameterization and the estimated conditions for the base of the MOs, the average DCmetal/silicate value for Earth, Mars, and the Moon can be predicted. The deep MO of Earth is predicted to cause the strongest depletion of its silicate carbon budget, closely followed by Mars with intermediate depth MO, and then the Moon with a shallow MO. We predict that the lunar mantle carbon budget, similar to that of the Earth’s present-day upper mantle, might have been set by equilibrium core-mantle fractionation in MO; whereas for Earth, later processes such as ingassing from a proto-atmosphere and late-stage accretion of volatile-rich material was necessary for delivery of carbon and other volatiles. Finally, the comparison of our measured and predicted value of DCmetal/silicate for terrestrial MO with similar constraints on DNmetal/silicate from the literature suggests that the apparent depletion of nitrogen relative to carbon for the bulk silicate Earth and the Earth’s upper mantle is unlikely to be caused by preferential partitioning of nitrogen to alloy melt during core formation.

Tungsten isotopic evolution during late-stage accretion: Constraints on Earth–Moon equilibration

We couple the results of N-body simulations of late-stage accretion (O'Brien et al., 2006) to a hafnium–tungsten (Hf–W) isotopic evolution code to investigate the evolution of planetary bodies in the inner solar system. Simulations can simultaneously produce planets having Earth- and Mars-like masses and Hf–W systematics by assuming that the tungsten partition coefficient decreases with increasing semi-major axis (e.g. due to increasing oxidation). Simulations assuming that Jupiter and Saturn occupy circular orbits are more successful at reproducing the Hf–W systematics than those assuming present-day Jupiter and Saturn orbits. To generate Earth-like tungsten anomalies, 30–80% of each impactor core is required to re-equilibrate with the target mantle. Some model outcomes yield a target and final impactor having similar (Earth- and Moon-like) tungsten anomalies. However, in no case can the inferred lunar Hf/W ratio be simultaneously matched. This result suggests that the Moon isotopically equilibrated with the Earth's mantle in the aftermath of the giant impact (cf. Pahlevan and Stevenson, 2007). Alternatively, either the dynamical models which show the Moon being derived primarily from the impactor mantle, or the accretion timescales obtained by the N-body simulations, are incorrect.

PROTOSTAR FORMATION IN SUPERSONIC FLOWS: GROWTH AND COLLAPSE OF SPHERICAL CORES

We present a unified model for molecular core formation and evolution, based on numerical simulations of converging, supersonic flows. Our model applies to star formation in GMCs dominated by large-scale turbulence, and contains four main stages: core building, core collapse, envelope infall, and late accretion. During the building stage, cores form out of dense, post-shock gas, and become increasingly centrally stratified as the mass grows over time. When the shock radius defining the core boundary exceeds $R\approx 4 a (4\pi G \rho_{mean})^{-1/2}$, where $a$ is the isothermal sound speed, a wave of collapse propagates from the edge to the center. During the building and collapse stages, density profiles can be fit by Bonnor-Ebert profiles with temperature 1.2 - 2.9 times the true value. As found previously for initially static equilibria, outside-in collapse leads to a Larson-Penston density profile $\rho \approx 8.86 a^2/(4 \pi G r^2)$. The third stage, consisting of an inside-out wave of gravitational rarefaction leading to $\rho\propto r^{-3/2}$, $v\propto r^{-1/2}$, is also similar to that for initially-static spheres, as originally described by Shu. We find that the collapse and infall stages have comparable duration, $\sim t_{ff}$, consistent with estimates for observed prestellar and protostellar (Class 0/I) cores. Core building takes longer, but does not produce high-contrast objects until shortly before collapse. The time to reach core collapse, and the core mass at collapse, decrease with increasing inflow Mach number. For all cases the accretion rate is $\gg a^3/G$ early on but sharply drops off; the final system mass depends on the duration of late-stage accretion, set by large-scale conditions in a cloud.

Archean plate tectonics, rise of Proterozoic supercontinentality and onset of regional, episodic stagnant-lid behavior

Intense devolatilization and chemical-density differentiation attended late-stage accretion of the primitive Earth; it lessened after crystallization of a magma mush ocean during continued cooling. By ∼ 4.3Ga, shallow seas were present, so surface temperatures had fallen far below the ∼ 1300, ∼ 1120, and ∼ 950°C low-pressure solidi of peridotite, basalt, and granite, respectively. At temperatures less than about half their solidi, such materials existed as lithosphere in the near-surface Hadean realm. Stagnant-lid convection probably did not occur because massive heat transfer necessitated vigorous crust–mantle overturn in the early, hot Earth. Instead, bottom-up mantle convection, including voluminous plume ascent, efficiently rid the planet of heat, but lessened over time. Plate thickening and broadening is reflected in the post-Hadean rock record. Stages of geologic evolution included: (a) 4.5–4.4Ga, early, chaotic magma mush ocean overturn and ephemeral lithospheric platelets; (b) 4.4–2.7Ga, growth of oceanic and diminutive continental plates, obliterated by return mantle flow prior to ∼ 4.0Ga, but the latter enlarging and gradually accumulating as largely submarine, sutured, sialic crust-capped lithospheric collages; (c) 2.7–1.0Ga, progressive assembly of old shields and younger orogenic belts into supercratonal plates characterized by continental freeboard, sedimentary differentiation, and episodic glaciation during transpolar drift, as well as onset of regionally, temporally limited stagnant-lid convection beneath supercontinents; (d) 1.0Ga-present, modern, laminar-flowing asthenospheric cells capped by giant, stately moving plates. Restriction of komatiitic lavas to the Archean, and of multicycle sediments, most ophiolite complexes ± alkaline igneous rocks, and high-pressure and ultrahigh-pressure metamorphic belts to progressively younger Proterozoic–Phanerozoic orogens reflects increasingly negative buoyancy of the cooler oceanic lithosphere. Attending supercontinent assembly, density instabilities of thickening oceanic plates increasingly began to dominate overturn of the suboceanic mantle as cold, top-down convection. Scales and dynamics of hot asthenospheric upwelling versus lithospheric foundering and asthenospheric return flow (bottom-up versus top-down) changed gradually over geologic time in response to planetary thermal relaxation.

A Decreased Probability of Habitable Planet Formation around Low‐Mass Stars

Smaller terrestrial planets (≲0.3 M⊕) are less likely to retain the substantial atmospheres and ongoing tectonic activity probably required to support life. A key element in determining whether sufficiently massive sustainably habitable planets can form is the availability of solid planet-forming material. We use dynamical simulations of terrestrial planet formation from planetary embryos and simple scaling arguments to explore the implications of correlations between terrestrial planet mass, disk mass, and the mass of the parent star. We assume that the protoplanetary disk mass scales with stellar mass as Mdisk ∝ fM, where f measures the relative disk mass and 1/2 < h < 2, so that disk mass decreases with decreasing stellar mass. We consider systems without Jovian planets, based on current models and observations for M stars. We assume the mass of a planet formed in some annulus of a disk with given parameters is proportional to the disk mass in that annulus and show with a suite of simulations of late-stage accretion that the adopted prescription is surprisingly accurate. Our results suggest that the fraction of systems with sufficient disk mass to form >0.3 M⊕ habitable planets decreases for low-mass stars for every realistic combination of parameters. This habitable fraction is small for stellar masses below a mass in the interval 0.5-0.8 M☉, depending on disk parameters, an interval that excludes most M stars. Radial mixing and therefore water delivery are inefficient in the lower mass disks commonly found around low-mass stars, such that terrestrial planets in the habitable zones of most low-mass stars are likely to be small and dry.

Accretion Efficiency during Planetary Collisions

We present the results of smoothed particle hydrodynamic simulations of collisions between two 0.10 M⊕ differentiated planetary embryos with impact dynamics that are thought to be common to the late stage of terrestrial planet formation. At low impact velocities (vimp/vesc < 1.5) and for direct collisions, the impacts are largely accretionary. Inelastic bouncing between embryos with varying degrees of erosion, followed by escape to infinity, is also a common outcome. For dynamical environments typical of most late-stage accretion models, we estimate that more than half of all collisions between like-sized planetary embryos do not result in accumulation into a larger embryo.

Fossil 26Al and 53Mn in the Asuka 881394 eucrite: evidence of the earliest crust on asteroid 4 Vesta

Asuka 881394 is a unique magnesian eucrite with pyroxenes that are Mg-rich like those of cumulate eucrites, but with a granulitic texture unlike the textures of cumulate eucrites. Plagioclase compositions are ∼An 98, and are even more calcic than those in cumulate eucrites. Pyroxene does not show pigeonite-to-orthopyroxene inversion textures, suggesting different crystallization conditions than those of cumulate eucrites. Mn–Cr isotopic analyses determined initial 53Mn/ 55Mn=(4.6±1.7)×10 −6 and initial ϵ( 53Cr) I=0.25±0.17 in A881394. This initial 53Mn abundance corresponds to a formation interval Δ t LEW=−6±2 Ma relative to the LEW86010 angrite, implying an ‘absolute’ age of 4564±2 Ma. Both the initial 53Mn abundance and the initial ϵ( 53Cr) I value for A881394 are identical to those previously determined for the HED parent body at the time of its differentiation. Al–Mg isotopic analyses determined initial 26Al/ 27Al=(1.18±0.14)×10 −6, from which a formation interval Δ t CAI=3.95±0.13 Ma is calculated relative to the canonical value 26Al/ 27Al=5×10 −5 for CAI. Combining this formation interval with a recently reported Pb–Pb age of 4567.2±0.6 Ma for CAI gives 4563.2±0.6 Ma as the age of A881394, in excellent agreement with the age based on the Mn–Cr formation interval. Alternatively, the 53Mn and 26Al formation intervals of A881394 allow the Mn–Cr and Al–Mg timescales to be intercalibrated, suggesting that an ‘absolute’ CAI age of 4568 Ma is most consistent with the 4558 Ma Pb–Pb age of LEW86010. The initial 26Al abundance existing in A881394 would have been insufficient to cause global melting in the HED parent body (probably asteroid 4 Vesta). Nevertheless, it could have been derived by radioactive decay over only ∼2 Ma from an abundance that would have been sufficient to cause global melting. The higher value of molar Mg/(Mg+Fe)=0.57 for A881394 than those of the ordinary (basaltic) eucrites (Mg/(Mg+Fe)=0.30–0.42) suggests additional factors may have been important for magma genesis on the parent body. If 26Al were the only heat source, partial melting would have been the major process in the interior of the parent body, and Mg/(Mg+Fe) would be lower in the melts than in the primordial source material. Late-stage accretion could have supplied relatively magnesian primordial material to the surface of the parent body, thereby increasing Mg/(Mg+Fe) in a shallow magma ocean from which A881394 crystallized, and also may have augmented 26Al heating. The granulitic texture of A881394 may have been produced during residence in the thin, earliest, crust, kept hot by the magma beneath it. If 26Al was, nevertheless, the major heat source for asteroidal melting, it may account for declining post-accretion heating of main belt asteroids with increasing heliocentric distance.

Metal–silicate partitioning and the incompatibility of S and Si during core formation

We have determined the partitioning of a number of siderophile and lithophile elements between liquid metal and liquid silicate phases under the high temperature, low pressure conditions at which core segregation is considered to have originated [2]. Oxygen fugacity was varied from 2 to 6 log units below the iron–wüstite (IW) buffer by increasing the Si content of the metallic phase from 100 ppm up to the 8% proposed by Allègre et al. [8] as being cosmochemically required in the earth's core. We find that Si and S, two of the principal candidates for the light element in the core are mutually exclusive in the metal phase. As oxygen fugacity is reduced, Si becomes increasingly siderophile and S increasingly lithophile. At 8% Si in the metal, all S, up to at least 1%, enters the coexisting ultramafic liquid. Thus, Si and S are incompatible during a single-stage low pressure core formation event. Mn and Cr, which are lithophile under oxidising conditions, become slightly siderophile if there is 8% Si in the metal, corresponding to an oxygen fugacity of about 6 log units below the iron–wüstite (IW) buffer. Ti remains lithophile even under these conditions, only becoming siderophile at about 7 log units below IW. These results demonstrate that the Mn and Cr contents of the core proposed in [8], coupled with lithophile behaviour of Ti, are consistent with core segregation under strongly reducing conditions. Such conditions could not generate a core containing 2% S, however, and the problem of the high contents of oxidised Fe, Ni, Co and other siderophile elements in the mantle would remain. The only plausible way of introducing S into the core is to add it as part of a late-stage accretion of oxidised material to the earth. A second, minor phase of core segregation under oxidising conditions would supply S, but no additional Si, to the core.