An experimental determination of chlorine isotope fractionation in acid systems and applications to volcanic fumaroles

Abstract

The δ 37Cl values of volcanic fumarole gases and bubbling springs were measured from the Central American and the Kurile arcs. Low temperature gas samples from the Central American arc have δ 37Cl values generally between −2 and 2‰, whereas high-temperature fumaroles (>100 °C) range from 4 to 12‰, with several outliers. This is in contrast to the high-temperature fumaroles from the Kurile island Kudryavy which have slightly positive δ 37Cl values, averaging 0.8‰ (±0.6, 1σ), and from our previous work on Izu and Mariana arc samples in which the δ 37Cl values of fumarole and ash samples are similar to each other and negative. Assuming that the source for the high-T Central American fumaroles has typical subduction δ 37Cl values (−2.5 to 1‰), then there must be a large Cl isotope fractionation in the near-surface fumarolic system. The most likely fractionation mechanism for the high δ 37Cl values is between Cl − aq − HCl( g), but published theoretical fractionation for this pair is only ∼1.5‰, insufficient to explain the large range of values observed in the fumaroles. Three experiments were undertaken in order to identify a process that could cause the wide range of δ 37Cl values observed in the high-temperature fumaroles. Results are the following: (1) A sub-boiling equilibration experiment between aqueous chloride and HCl gas had 1000 ln α HCl ( g ) - Cl - = 1.4 to 1.8 ‰ , in agreement with the theoretical calculations. (2) Evaporation of HCl( g) from hydrochloric acid at room temperature had fractionation in the opposite sense, with a 1000 ln α HCl ( g ) - Cl - = - 3.92 ‰ . (3) A ‘synthetic fumarole’ gave large positive fractionations up to 9‰, with 37Cl strongly partitioned into the vapor phase. The ‘fumarole’ experiments were made by bubbling dry air through boiling hydrochloric acid in an Erlenmeyer flask, and collecting the evolved HCl( g) in a second ‘downstream’ flask filled with distilled water. This extreme enrichment is likely due to a distillation process in which 37Cl-enriched HCl( g) is stripped from the hydrochloric acid followed by a significant fraction of the light HCl( g) redissolving into the constantly condensing water vapor on the walls of the first flask. This distillation experiment creates a Cl isotope fractionation that is the same order of magnitude as observed in the high-temperature fumaroles in Central America. These results suggest that there must be a H 2O liquid–vapor region in the sub-surface fumarole conduit where light Cl is stripped from the HCl gas as it passes through the fumarole. Similar 37Cl enrichments are expected in fossil epithermal boiling systems.