Pierre Agrinier

O and H stable isotope compositions of oceanic crust and upper mantle rocks exposed in the Hess Deep near the Galapagos Triple Junction

Abstract A large variety of rocks, consisting of basalts, dolerites, gabbros and ultramafics, are exposed on the Intra-rift ridge in the Hess Deep, representing a section of the oceanic crust formed at the East Pacific Ridge [1]. We conducted a stable isotope study of O and H to assess the nature of the interaction between seawater and the oceanic crust in this region. Fresh basalt and dolerite samples have magmatic MORB values (δ18O ≈ 6.0, −105 ≤ δD ≤ −70) and low H2O+ content ( −57) and H2O+ contents (> 1.3%). Most gabbros have δ18O ≈ 5.6, essentially identical to their initial magmatic values. Their δD values are generally higher than −57, suggesting that they have reacted with some seawater-derived fluids, as is also suggested by their mineralogy. In gabbros, displaying little or no visual evidence of alteration of their plagioclases and pyroxenes, Δ18Oplagioclase-pyroxene is small (≈ 0.3) and consistent with magmatic values. The calcic (an > 50) plagioclases have magmatic δ18O (≈ 5.8). Albitized, prehnitized plagioclases have variable δ18O values (4–8). It also appears that the 18O-modified gabbros are those in which plagioclases are transformed into albite and/or prehnite, which are characteristic of greenschist facies conditions. The lack of calcic (an > 50) plagioclases depleted in 18O and hornblende suggests that most of the gabbros did not react extensively under amphibolite facies conditions with seawater-derived fluids (T ≥ 400 °C) or that they were subsequently altered to greenschist facies gabbros. The serpentinized peridotites were depleted in 18O (3.3–4.9). In one sample, the Δ18Oserpentine-magnetite (≈ 5.7) corresponds to a temperature of about 325 ± 50°C. These values are compatible with serpentinization by seawater-derived fluids at temperatures between 200° and 350°C. Another sample has a higher δ18O value of 10, corresponding to a serpentinization temperature of less than 50°C. The 18O mass balance between seawater and the oceanic crust seems to have attained steady-state where enrichment in 18O of metabasalts is balanced by the depletion in 18O of metadolerites. The contribution of the gabbros is minor. The depletion in 18O of the serpentinized peridotites is not taken into account since we believe that the high-temperature serpentinization is characteristic of the Hess Deep and cannot be extrapolated to the normal oceanic crust.

Fast back-reactions of shock-released CO 2 from carbonates: An experimental approach

This work aims at investigating the processes leading to the liberation of CO 2 and SO2/SO3 in the atmosphere after large meteorite impacts into sediments. Firstly, we review reactions and thermodynamic conditions to produce CO2 from carbonates and SO2/SO3 from sulfates. We show that decomposition of the carbonates and sulfates only occurs during shock pressure release. Secondly, we examine mineralogical and chemical data of natural impact breccias where pure CaO (lime) is always lacking and where secondary carbonates and probably sulfates occur abundant. This observation evidences the importance of back-reactions of CO2 and SO2/SO3 with the initially produced CaO. Third, we explore the kinetics and thermodynamics of the reactions involving CaO and CO2. We have performed 32 degassing and back-reaction experiments with fine-grained, chemically precipitated calcite, and with coarse-grained natural calcite, dolomite, and magnesite. Experiments with calcite confirm that residual CaO is highly reactive in the presence of CO2 in the 573-973 K interval: within less than 200 s, some 40 to 80% of CaO has back-reacted into CaCO3. These high reaction rates suggest that much of the impact produced CO2 may be highly transient. Scanning electron microscope observations show that these high reaction rates are enhanced by the exceptionally porous structure of the residual CaO. The kinetics of the CaO 1 CO2 reaction are explained by a gas-solid reaction model, in which the reaction rates are controlled by gas mass transfer through the porous CaO, the CO2-CaO surface interactions, and the diffusion of CO2 through CaCO3. Similar experiments conducted with dolomite and magnesite show that residual Mg-oxides do not react significantly at the 1000 s time scale and may, therefore, survive as witness of degassing in impact breccias. Published kinetic modeling of SO2/SO3 back-reactions with hot CaO to CaSO4 indicates typical conversion rates of around 50% after 1200 s. Hence back-reactions play also a crucial role in limiting the total amount of sulfur oxides released by an impact event into the Earths atmosphere and stratosphere. At low tempera- tures, residual CaO should react with water to yield Ca(OH)2 (another very efficient CO 2 pump), or dissolve in natural waters strongly increasing the pH. This pH effect is globally compensated by the acid species (H2CO3 ,H 2SO4) produced from liberated CO2 and SO2/SO3. Our experimental data, and the assessment of existing literature indicate that the amount of chemically active gases that have been released into the atmosphere by the Chicxulub impact event are most likely overestimated. Copyright

Shock recovery experiments on dolomite and thermodynamical calculations of impact induced decarbonation

We have studied experimentally shocked dolomites and calcites by scanning electron microscopy (SEM), analytical transmission electron microscopy (ATEM), X ray diffractometry (XRD) using Rietveld refinements, and mass spectrometry analysis of the abundance ratios of the stable isotopes of carbon and oxygen. The shock-recovery experiments have been perfomed by the multiple reverberation technique on natural dolomite rocks at 60 GPa, using steel devices and highexplosive driver-flyer plates as plane shock wave generators. Modified assemblies with grooves and holes were built in order to facilitate the escape of CO 2 , in case the conditions of breakdown of the carbonates into oxides and CO 2 would be reached during the shock or postshock history of the samples. In contrast with the results from previous studies, almost no evidence for outgassing, expressed, for example, by the identification of CaO or MgO, could be observed. Consequently, no isotope fractionation occured in the shocked samples. This result is consistent with the calculations of peak-shock and postshock temperatures, as well as with the examination of outgassing conditions of carbonates, calculated in this study up to 80 GPa. We have shown that in a direct shock, outgassing in air of nonporous dolomites and calcites should occur in impacts at 55-65 GPa and 35-45 GPa, respectively. The effect of porosity, which strongly lowers these values, has been estimated. The experimental setup for shock-recovery experiments is shown to be an important parameter : reverberated shocks lead to lower peak-shock and postshock temperatures than direct shocks at the same pressure ; differences among experimental setups might explain part of the discrepancies between previous studies. In this study, the only surviving shock-induced phenomenon is pulverization leading to grain sizes smaller than in the starting material. The decrease in grain size has been quantified via a structure refinement by Rietveld analysis of X ray powder patterns, which also allows the estimation of the lattice strains. A future pressure scale of shock effects in carbonates could probably be based on such parameters.

A SEM-ATEM and stable isotope study of carbonates from the Haughton impact crater, Canada

Abstract Highly and intermediately shocked carbonate-rich fragments of the allochtonous polymict breccia from the Haughton impact crater (Canada) were studied by Scanning Electron Microscopy (SEM), Analytical Transmission Electron Microscopy (ATEM) and analyses of carbon and oxygen stable isotopes ( δ 13 C and δ 18 O). In areas subjected to severe shock conditions, carbonates represent only about 10 vol% of the shocked samples and they are located in holes and fractures within a matrix of SiO 2 -rich glass. Shock features are absent in these crystals. High-temperature reactions have occurred between molten silicates and carbonates, producing Ca Mg-rich glasses, or crystalline phases such as augite and larnite (Ca 2 SiO 4 ). The carbonates are dominated by calcite and they generally have significantly positive δ 13 C, ranging up to +9‰, with a weighted average value of +1.75‰. Their δ 18 O values range between +15‰ and +20‰ and they are about 5‰ lower than in unshocked reference sediments, a trend consistent with that resulting from silicate-carbonate reactions. The microstructures of the carbonates suggest that they did not undergo shock conditions but, instead, were produced by back-reactions between impact-released CO 2 and highly reactive residual oxides. Such a process would introduce isotope fractionations, which might explain the positive δ 13 C values observed. A simple kinetic fractionation model involving a Rayleigh distillation process is used to estimate the CO 2 fraction actually lost from the carbonates. It appears that this fraction is related to the amount of high-temperature carbonate-silicate reactions. Moderately shocked fragments from other areas of the polymict breccia consist of 40–81 vol% carbonates. Their δ 13 C values lie in the range of unshocked reference sediments between −2‰ and −4‰, whereas their δ 18 O values are by about 5‰ lower than in the unshocked equivalents. No evidence for important decarbonatization is observed from 13 C, and 18 O is again buffered by isotope exchange reactions between molten silicates and carbonate crystals producing Ca and Mg enriched SiO 2 glass and Ca Mg silicate crystals such as monoclinic pigeonite, which is indicative of fast cooling. This study indicates that significant evidence for outgassing is limited to a narrow zone in the centre of the crater, where peak shock pressures reached 50–60 GPa. Moreover, we suggest that, within this area, a large fraction of the shock-produced gas recombines with the highly reactive residual oxides and, in consequence, that such back-reactions might be a general mechanism for retaining impact-produced volatiles during impact events.

Les cratères d'impacts: principaux effets de choc dans les roches et minéraux

Abstract The basic principles of the physics of shock waves are summarised, showing how toxic pressures, shock and pot-shock temperatures, and shock durations can be estimated in the case of large meteorite impacts on Earth. In a second part, the pertinence of laboratory high-pressure dynamic experiments for simulating large meteorite impact events and for calibrating their physical conditions is discussed. It is concluded that most shock features are common to natural and laboratory shocks, although the lifetime of experimental shocked states is shorter by several orders of magnitude. Then, a review is, made of the major shock effects observed in minerals and rocks. Quartz has been, by far, the most extensively studied shocked mineral. Particularly, planar deformation features (PDFs), interpreted as resulting from relaxations at the shock front, are unambiguous shock indicators, for shock pressures approximately between 15 and 35 GPa. At higher pressure the formation of high-pressure polymorphs of SiO 2 in shocked quartz is also discussed. Shock effects in some other selected minerals, although less extensively studied, are also reviewed, with special emphasis on the discovery of diamonds at impact sites and of all the high-pressure polymorphs of olivines and pyroxenes, including silicate perovskite, in shocked meteorites. Finally, the controversial links between large impacts and major environmental effects are discussed in a fourth part.