Chlorine isotopic composition in seafloor serpentinites and high-pressure metaperidotites. Insights into oceanic serpentinization and subduction processes

Bulk-rock chlorine content and isotopic composition (δ³⁷Cl) were determined in oceanic serpentinites, high-pressure metaperidotites and metasediments in order to gain constraints on the global chlorine cycle associated with hydrothermal alteration and subduction of oceanic lithosphere. The distribution of insoluble chlorine in oceanic serpentinites was also investigated by electron microprobe. The hydrothermally-altered ultramafic samples were dredged along the South West Indian Ridge and the Mid-Atlantic Ridge. The high-pressure metamorphic samples were collected in the Western Alps: metaperidotites in the Erro-Tobbio unit and metasediments in the Schistes Lustrés nappe.Oceanic serpentinites show relatively large variations of bulk-rock Cl contents and δ³⁷Cl values with mean values of 1105 ± 596 ppm and −0.7 ± 0.4‰, respectively (n = 8; 1σ). Serpentines formed after olivine (meshes) show lower Cl content than those formed after orthopyroxene (bastites). In bastites of two different samples, Cl is positively correlated with Al₂O₃ and negatively correlated with SiO₂. These relationships are interpreted as reflecting preferential Cl-incorporation into the bastite structure distorted by Al (substituted for Si) rather than different alteration conditions between olivine and orthopyroxene minerals. High-pressure metaperidotites display relatively homogeneous Cl contents and δ³⁷Cl values with mean values of 467 ± 88 ppm and −1.4 ± 0.1‰, respectively (n = 7; 1σ). A macroscopic high-pressure olivine-bearing vein, formed from partial devolatilization of serpentinites at ∼2.5 GPa and 500-600 °C, shows a Cl content and a δ³⁷Cl value of 603 ppm and −1.6‰, respectively. Metasediments (n = 2) show low whole-rock Cl contents (<15 ppm Cl) that did not allow Cl isotope analyses to be obtained.The range of negative δ³⁷Cl values observed in oceanic serpentinites is likely to result from water-rock interaction with fluids that have negative δ³⁷Cl values. The homogeneity of δ³⁷Cl values from the high-pressure olivine-bearing vein and the metaperidotite samples implies that progressive loss of Cl inherited from oceanic alteration throughout subduction did not significantly fractionate Cl isotopes. Chlorine recycled in subduction zones via metaperidotites should thus show a range of δ³⁷Cl values similar to the range found in oceanic serpentinized peridotites.     

Origin and evolution of fluids from mud volcanoes in the Barbados accretionary complex

A large collection of fluids (54 interstitial fluids and four expelled fluids) were sampled at the Manon site, at the outer edge of the Barbados accretionary complex. These warm fluids (up to 20°C) are expelled by sub-marine (5000 mbsl) mud volcanoes consisting of diapirs (unchanneled flow) and diatremes (channeled).Chlorine stable isotope ratios of these fluids were measured by IRMS with a reproducibility of ± 0.05‰ (1σ) versus SMOC (Standard Mean Ocean Chloride).A large range of δ³⁷Cl between −5.3‰ and +0.1‰ is observed. Data from each volcanic structure describe a mixing between seawater and a low-δ³⁷Cl fluid. The whole set of data is interpreted as the result of a mixing between two deep components and seawater. The two deep fluids are chemically distinct (e.g., in Ca, Mg, K, Li, Sr and Br contents and Br/Cl ratio). They display low and significantly different ⁸⁷Sr/⁸⁶Sr ratios (0.707790 and 0.707892, respectively) and δ³⁷Cl values (−4.51 and −5.24‰, respectively).Physicochemical processes such as mineralogical transformation, diffusion, compaction or ion filtration are known to fractionate chlorine stable isotopes and can produce fluids with negative δ³⁷Cl values. Ion filtration due to sediment compaction appears to be the more likely process to explain the negative δ³⁷Cl values observed at the Manon site. A model for the generation of these signatures is proposed where a residual negative δ³⁷Cl fluid reservoir is created at the bottom of the prism or the sediment pile. Further compaction/fracturing and/or dewatering of the slab may flush out these fluids and focus them towards the décollement zone. Mixing between the fluids and ultimately with seawater and water released during gas hydrate destabilizations may explain the data set within the individual cores and between the different structures.     

Overcoming problems of density and thickness measurements in FTIR volatile determinations: a spectroscopic approach

The Beer–Lambert law is traditionally used to determine water and carbon concentrations in glasses from their infrared (IR) spectra. In practice, this method requires estimation of the thickness and density of the glass as well as the calibration of the molecular absorptivities of the species concerned. All of these parameters can be sources of practical difficulties and analytical uncertainty. These weaknesses in the application of the Beer–Lambert law have been overcome by an empirical analysis of the infrared spectra. Using a set of 292 spectra obtained on 113 natural and experimental tholeiitic glasses (SiO₂ 48.5–51 wt%; water contents 0–4000 ppm H₂O), it can be shown that the thickness–density (ρ d) product of a glass sample can be directly and reliably inferred from its IR spectrum. This allows the Beer–Lambert law to be rewritten. The new form no longer requires thickness or density estimations to determine volatile contents. Moreover, if needed, the thickness of the glass slab can also be accurately determined from the IR spectra. This new method is developed for quantitative determination of water concentrations in MORB glasses but can also be applied to any minor species (carbon, sulfur, etc.) provided it is active in the IR domain and that a suitable independent frequency of IR absorption can be identified. Precision is about 60 ppm H₂O on O–H⁻ contents. This method, tested on natural and experimental MORB-type glasses, can be applied to any chemical composition provided a set of reference spectra is available. Received: 16 September 1999 / Accepted: 18 February 2000     

Chlorine transfer out of a very low permeability clay sequence (Paris Basin, France): Cl and Cl evidence

The marine Callovo-Oxfordian clay formation is found at a depth around 410m in the eastern part of the Paris Basin (France). It is a very low permeability formation investigated by the French agency for nuclear waste management (ANDRA) to study the feasibility of a radioactive waste disposal. Examining hydrogeological and geochemical characteristics of the clay sequence may test confinement properties of this formation. This study uses chlorine isotopes to investigate long-term transport processes which may carry chemical elements out of the clay layer to the surrounding rocks. Detailed chlorine concentration and δ³⁷Cl depth profiles are examined using pore waters and aquifer waters sampled in the clay formation and its surrounding aquifers (the Dogger at the bottom and the Oxfordian/Kimmeridgian/Tithonian unit at the top). They are discussed in terms of chlorine budget and hydrogeological processes.Clay pore waters and aquifer waters show strong chlorine concentration depletion (<3000 mg/L) relative to the original marine interstitial water (∼19000 mg/L). This probably results from an early dilution by meteoric water in limestones (as also indicated by oxygen and hydrogen isotopes).A steep Cl-concentration gradient from the Dogger at ∼500m in depth (∼2500 mg/L) to the Oxfordian/Kimmeridgian/Tithonian aquifer near the surface (≈ 10 mg/L) is associated to a ‘v-shaped’ profile of the δ³⁷Cl values. Modelling Cl transport shows that a hydrodynamic dispersion process explains Cl concentration and δ³⁷Cl profiles in Oxfordian Limestone. This process implies a mean upward flux of chloride in the 2.6 10⁻⁸-8.2 10⁻⁸ mole/m²/yr range from the clay formation towards upper limestones where a westward advective flow disperses the chloride. The modelling and knowledge of underground water transfer suggest a maximum effective Cl-hydrodynamic vertical dispersion coefficient (= vertical Cl-transport coefficient) of ∼7.6 10⁻¹⁰ m²/s.Chlorine transfer through the Callovo-Oxfordian clay, since deposition 160My ago, can be mainly described by the interplay of an early dilution and a later hydrodynamic dispersion event which has apparently erased most of the isotopic effects of diagenetic events (such as early diffusion, ion filtration etc.).