V11E-01
The Neodymium 142 Puzzle Solved: Hadean Isotopic Evolution in the Protosolar Nebula and Early Earth
Since the discovery of a difference in 142Nd/144Nd between the Earth and chondrites (Boyet and Carlson, 2005), a puzzle has been created for understanding Earth evolution and composition: do we have a chondritic bulk Earth composition, necessitating a hidden enriched reservoir in its interior, or is the Earth non-chondritic? Are we losing the historical reference frame for the composition of the Earth? We propose a hypothesis that explains the observed difference in 142Nd/144Nd as a result of different initial 142Nd/144Nd ratios for the Earth and chondrites, with the Earth sampling the solar nebula later than chondrites, thus giving additional time for 146Sm decay. This hypothesis implies fractionation between Sm and Nd during condensation of solid materials from a nebula with higher Sm/Nd ratio. Several lines of evidence in meteorites as well as in experiments have indeed shown that Sm is more volatile under reducing conditions; this results in significant reduction in Sm/Nd ratio in condensates relative to the nebula. Furthermore, observations of nebulae around young stars have shown that they indeed possess reducing conditions (high C/O ratios). Observations also indicate that the process of formation of solid materials is not a simple one-way condensation, but a complex sequence of several steps of evaporation-condensation processes that could last for several million years. In addition, 129Xe has shown that the Earth accreted its final materials 120 m.y. later than meteorites, providing support for our hypothesis. We interpret the observation of 142Nd/144Nd ratios in Archean (3.8 Ga) rocks, showing positive 142ε regardless of acidic or basaltic composition, as witnesses of late accreting materials with high 142Nd/144Nd from the nebula. These anomalies vanish with time, because they are destroyed by the geodynamic cycle.
V11E-02
Timing of Formation of the Earth-Moon System
It is generally agreed that proto-Earth had accreted from planetesimals of broadly chondritic composition by 4.55 Ga. It is also widely believed that the Earth-Moon system formed following the subsequent collision with a Mars-sized-object. Such an event would surely have vaporized a large portion of both impactor and target and melted the rest of the combined system. However, the timing of this hypothesized event and the duration of crystallization of the molten outer layers of both bodies are controversial. The timing of terrestrial core formation (which would place an upper bound on the age of formation of the metal-poor Moon and thus collision) is not well known. Hf-W model ages range from 20 to >100 m.y., depending on assumed value of Hf/W in the silicate Earth or Moon, but recent reports favor the higher end of this range. 142,143Nd/144Nd data for lunar basalts have been interpreted to indicate that the lunar magma ocean (LMO) was still active at 4.35-4.31 Ga, although such relatively youthful ages could instead reflect isotopic resetting or source heterogeneity. Interpretation of zircon Lu-Hf data, however, is less sensitive to underlying assumptions that limit the aforementioned estimates. We have recovered exotic zircons in the age range 4.35 to 3.91 Ga from both terrestrial (Jack Hills quartzites) and lunar (Apollo 14 breccias) samples and undertaken laser ablation Lu-Hf analyses to obtain initial 176Hf/177Hf ratios (Hfo). Terrestrial zircons (n=221) show dominantly negative εHf(T) throughout the Hadean and yield an essentially uniform spectrum of mantle extraction ages between 4.5-4.2 Ga. Four robust analyses yield Hfo close to the solar system initial indicate source sequestration in an ultra-low 176Lu/177Hf environment by 4.52±0.02 Ga (using the most recent chondritic Lu-Hf parameters). The lunar zircons crystallized from highly incompatible-trace-element enriched (KREEP) sources that evolved during differentiation of the LMO. Nearly half of the model ages calculated assuming a source 176Lu/177Hf of 0.014 (derived from the overall data trend but consistent with independent estimates) cluster at 4.48±0.03 Ga which we interpret to be a lower bound on the last stage of LMO crystallization. We thus suggest that collisional events leading to the formation of the Earth-Moon system and the solidification of the molten outer layers of both bodies may have been complete within 70 m.y. of CAI formation.
V11E-03
Scaling Laws for Heat Partitioning in Terrestrial Planets During Core Formation
Core formation is the first major differentiation event that determines the initial conditions from which terrestrial planets have evolved until present. The separation of metal and silicates in terrestrial planets is likely to generate substantially large amounts of heat by conversion of potential energy into thermal energy via viscous heating. This gravitational heat release for an Earth or a Mars-like planet represents a temperature increase of few hundreds to few thousands Kelvin. While estimating the total amount of gravitational heating in a planet is relatively straightforward, determining its partitioning between the silicate and metallic part is not a trivial task. We thus investigate dynamically the heat partitioning between metal and silicate during core formation by negative diapirism. We model numerically the sinking of iron-rich diapirs through a viscous silicate mantle, in spherical axisymmetric geometry. We carried a parameter study in which shear heating as well as several viscous rheologies were considered and systematically varied. We developed a simple semi-analytical model that captures the essential physics and we subsequently use it to derive general scaling laws for the time evolution of heat distribution between the iron-diapirs and the silicate material. These scaling laws are then used to determine the heat distribution within terrestrial planets during their growth and early differentiation.
V11E-04
Early depletion of the mantle as evidence of ultra-deep melting
In the first 100My of Earth's history, major magmatic events differentiated the mantle into several silicate reservoirs, one of which could be unsampled at the surface, and therefore "hidden." Evidence for this scenario comes from the study of 142Nd in early Archean rocks that preserve fingerprints of early mantle depletion. A candidate for the complementary unsampled reservoir is a crystallizing basal magma ocean (BMO) at the bottom of the Earth's mantle (Labrosse et al., Nature 2007), whose remnants are seismically observed as mushy pockets above the core-mantle boundary. We present models in which the 142Nd, 143Nd and 176Hf signatures measured in the oldest rocks arise from sampling of crystals produced upon cooling of the BMO. The decoupling between Nd and Hf isotopes in early Archean rocks is explained by crystallization of Mg-perovskite from the BMO at very high pressures. Sampling of the crystals stopped when they became too dense to be entrained and mixed upward into the solid mantle by convection. Our model suggests that the depletion trend of the early Archean arises from sampling of witnesses to basal differentiation.
V11E-05 INVITED
Age and Nature of the Protocrust of the Jack Hills Zircon Host Rock
A fundamental question about Earth evolution is to understand when growth of continental crust began. Constraining this event will help date the timing of onset of plate tectonics and thus ultimately the origin of life. We have addressed this problem by Pb-Pb dating of single Hadean zircons from Jack Hills in Western Australia combined with analysis of their Hf isotopic compositions. We found an average age for the Hadean population of the Jack Hills zircons analyzed here of 4.1±0.1 Ga suggesting that the host rock of the Jack Hills zircons formed as a single event rather than a succession of pulses. Most of the grains have Hf isotopic compositions less radiogenic than both chondrites and the depleted mantle at that time indicating that the host rock of the Jack Hills zircons derived from an enriched reservoir, or source rock, with sub- chondritic Lu/Hf. Monte Carlo modeling gives an age for this source rock of 4.35 Ga and a 176Lu/177Hf ratio of 0.05. Such a low Lu/Hf ratio does not match either modern oceanic crust, island arc rocks, or subduction-zone type granites, but better fits Archean TTG suites, which themselves in turn may have been produced by melting of the last remains of the terrestrial magma ocean (hydrous KREEPy basalt). If this proposition is correct, then whether plate tectonics had started by the time the Jack Hills zircons formed at ~4.1 Ga hinges critically around the understanding of the processes that produced TTGs in the first place. If, as suggested by recent literature, TTGs are not subduction zone magmas, the Jack Hills zircons may not have been witnesses to early plate tectonics. If, however, TTGs do represent magmas formed at plate margins, the Jack Hills zircons become the oldest manifestation of plate tectonics operating under some form at ~4.35 Ga. Alternatively, the source of the host rock of the Jack Hills zircons could have been a partial melt of KREEPy material that formed an enriched oceanic crust that subducted and remelted a few 100's of My later.
V11E-06 INVITED
Generation and Reworking of Archaean and Hadean Crust
Combined Hf and O isotopes in well-dated zircons are increasingly used to investigate the age of the crustal source rocks of detrital and inherited zircons. O isotopes are used to screen out samples that may have a sediment contribution in the parental magma, since sediments yield hybrid model ages that are difficult to interpret. Mafic and granitic rocks also have different Lu/Hf ratios, and so in principle the Hf isotope ratios of zircons can be used to investigate the broad composition of the average crust. The unradiogenic Hf isotope compositions of the Jack Hills zircons from Western Australia indicate the existence of enriched (crustal) reservoirs by at least 4.3 Ga (Y. Amelin et al., 1998, Nature v. 399, p. 252- 255; T. M. Harrison et al., 2005, Science, v. 310, p. 1947-1950). We report in situ Hf isotope analyses of the Jack Hills zircons in which the Pb isotope age information is measured concurrently with the Hf isotope data. The simple data arrays provide clear evidence for Earth differentiation at 4.5 Ga, with the production of both continental crust-like material and a mafic crustal reservoir with higher Lu/Hf. The continued resampling of this reservoir over at least 1.5 Ga argues for a substantial stabilised volume of mafic crust, and, in tandem with oxygen isotope data, the existence of Hadean continents. Zircons remain poor windows into the upper mantle. We therefore investigate Nd isotopes in well-dated titanites; they have closure temperatures for Pb in the range 600-750oC and they can retain cores with distinct age and REE chemistry to subsequent rim overgrowths. Nd isotopes offer a complementary approach to Hf in zircon that can be used to construct the both depleted mantle evolution and crustal growth curves.
V11E-07
Exploring for early bombardments on Earth from pre-3.85 Fa thermal effects recorded in Hadean zircons - a status report
We report on our progress with high-resolution ion microprobe U-Th-Pb depth profiles and Ti+REEs spot analysis which show that subsequent to their crystallization in melts under typical crustal conditions on Earth, some Hadean (pre-3.85 Ga) zircons record common age domains with unusual chemical and isotopic characteristics consistent with a high-temperature (possibly impact) origin. We have found evidence for later overprints caused by intense thermal alteration between 3.94-3.97 Ga in six of eight studied grains but no evidence for older events. These findings alert us to two fundamental things we did not know before about the probiotic potential of the Earth in the earliest solar system: (i) that the bombardment epoch did not result in complete 'Doomsday' scale destruction of the Earth's crust since the Moon-forming event at ca. 4.5 Ga; and (ii) age constraints on both sides of the ther-mally altered 3.94-3.97 Ga zircon domains are very good and so far our data show that no detectable thermal events are recorded by the zircons before ~3.97 Ga up to about 4.3 Ga. This observation is consistent with the output of new classes of dynamical models that successfully re-create the decay of impactor populations in the early solar system as recorded on the Moon and in meteorites.
V11E-08
Thermal State of the Lithosphere During Late Heavy Bombardment: Implications for Early Life
We model thermal effects of impacts on the terrestrial lithosphere during the period of Late Heavy Bombardment (LHB), a putative ~100 Ma epoch of sharply elevated impactor flux that reached a maximum at ca. 3.9 Ga. The goals of this work include estimating the degree to which the crust was molten or thermally metamorphosed during the LHB and evaluating habitability during this time period. We created a stochastic cratering model which populates all or part of the Earth's surface with craters within a probability field of constraints derived from the lunar cratering record, the size/frequency distribution of the asteroid belt, and dynamical models. For each crater in the model, a temperature field was calculated using analytical expressions for shock-deposited heat and central uplift. The resulting thermal anomaly was then introduced into a 3-dimensional model of the lithosphere, and allowed to cool by conduction in the subsurface and radiation/convection at the atmosphere interface. Parameters tested in the model include the duration, mass flux, and average impact velocity during the LHB, mean lithospheric thickness, lithospheric composition, and the presence or absence of oceans. We also assessed habitability by monitoring habitable volumes for mesophile (~20-50° C), thermophile (~50-80° C), and hyperthermophile (~80-110° C) microbial life in what we term the "geophysical habitable zone"; the volume of inhabited crust within ~4 km of the surface. Results of this work indicate that most of the crust was not melted or thermally metamorphosed to a significant degree under any reasonable scenario evaluated. Smaller impactors (1-10 km) were as important as gigantic basin formers (100+ km) in terms of sterilizing the habitable zone in the near-surface due to their far greater numbers (~170,000 impactors in the 1-10 km diameter range versus ~30 impactors in the 100+ km diameter range). However, large basin-forming craters are nonetheless more thermally and biologically significant because they take a far longer time to cool and drive long-lived hydrothermal systems (~ 106-107 years). Habitable conditions in the near- surface are re-established quickly after crater formation (~ 104-105 years for craters in the 100-1,000 km range), particularly if water is present. Although mesophile habitable volume decreases and hyperthermophile habitable volume increases during the LHB, the total habitable volume does not significantly change in most scenarios. The habitable volume in active hydrothermal systems always increases as the LHB progresses, but hydrothermal environments typically constitute a relatively small fraction (~1%) of the total habitable volume. We find that the LHB was insufficient to extinguish microbial life in the crust under any scenario explored in our model.