δ37Cl of Mid-Ocean Ridge Vent Fluids Determined by a new SIMS Method for Stable Chlorine Isotope Ratio Measurements

A method has been developed for the direct determination of δ³⁷Cl in natural fluid samples or rock leachates (pyrohydrolysis products) using Secondary Ion Mass Spectrometry (SIMS). Samples as small as 1 μl (<2 μg Cl) are simply dried by evaporation and the residual salts are then admixed with 1 mg of graphite powder and pressed into a small (1mm diameter) pellet amenable to SIMS analysis. Analyses are performed with a large format high-transmission, high-resolution ion microprobe -- the IMS 1270. Mass resolving powers of greater than 5000 are used to exclude isobaric interferences on ³⁵Cl and ³⁷Cl, producing an accurate and reproducible measurement of δ³⁷Cl. Nine analyses of NIST Cl isotope standard 975a yield an external reproducibility of 0.5 ‰ (2σ ). Repeat analyses of samples are reproducible within 1 ‰ . First δ³⁷Cl data for mid-ocean ridge hydrothermal vent fluids from three sites at EPR 9°N and the Logatchev site (MAR 15°N) have been collected. End member δ³⁷Cl compositions for the EPR fluids are between +6.5 and +7.1 ‰ , whereas that of the Logatchev fluid is +4.6 ‰ . Together with pore waters from accretionary prisms, which are depleted in ³⁷Cl (-2.0 to -7.7 ‰ ; Ransom et al., Geology, 23, 715-718, 1995), seawater-derived fluids in the marine environment span range in δ³⁷Cl of 15 ‰ . This variability is remarkably large when compared to >100 analyses of continental waters (formation and oil-field waters, fresh waters, brines, etc.) that cluster around 0 ‰ with a maximum variation of only 5 ‰ . Two observations suggest that the ³⁷Cl enriched nature of the vent fluids is not related to phase separation. (1) Laboratory experiments indicate that the Δ³⁷Cl_{(vapor-brine)} associated with super-critical phase separation of seawater between 420 and 450°C is small (-0.6 to 0.2 ‰ ; Magenheim, PhD Thesis, UCSD, 1995). (2) Conjugate vapor-brine pairs of boiling sampled in 1991 and 1994 at F vent (Von Damm et al. EPSL, 149, 101-111, 1997) have basically identical δ³⁷Cl values indicating that phase separation (sub- or super-critical) does not significantly fractionate chlorine isotopes. We suggest, therefore, that the heavy Cl isotope signature of the fluids is a result of seawater-rock interaction and/or mineral precipitation rather than phase separation of seawater. However, the specific mechanisms responsible for this enrichment are not yet understood. The Cl isotope signatures of hydrothermally altered gabbros and sheeted dike rocks (δ³⁷Cl: 0.4-3.4 ‰ ) tend to be heavier than seawater and it has been suggested that this indicates preferential uptake of ³⁷Cl by amphibole (Magenheim et al., EPSL 131, 427-432, 1995). Amphibolitization can therefore not account for the development of the ³⁷Cl enrichment of the fluids. Although the exact mechanisms of Cl isotope fractionation are not yet defined, our data suggest that Cl may not be a strictly conservative component in MOR hydrothermal systems.