Relationships between Na-Ca exchange, reaction temperature, and Sr isotopes in deep-sea hydrothermal fluids

Seawater circulating through ocean crust results in fluid-mineral exchange over a broad temperature range, and produces significant chemical fluxes to and from the oceans. Hydrothermal circulation is a sink for seawater Mg, primarily balanced by release of rock-derived Ca into the fluids. Once Mg is removed from the fluid, further charge-balanced cation exchange can occur, but it must involve mainly Ca and Na. When normalized to seawater chlorinity, [Na] and [Ca] in a global dataset of hydrothermal fluids exhibit a strong inverse correlation over a large concentration range ([Ca] 10-60 mmol; [Na] 500-400 mmol). The trend is offset from seawater composition because Ca initially increases independently of Na to compensate Mg loss. Reactive transport modeling suggests departures from seawater [Na] in the fluids can be toward higher or lower values depending on whether the secondary phase assemblage is relatively calcic (epidote and amphibole) or sodic (albite), and this is mostly related to temperature. Strontium concentrations in vent fluids show a positive and negative correlation with [Ca] and [Na], respectively. Although many fluids have rock-dominated ⁸⁷Sr/⁸⁶Sr, the lowest values (< 0.7035) correspond to fluids with the lowest [Ca]/[Na] and [Sr]/[Na] ratios, where [Na] is enriched and [Sr] is depleted relative to seawater concentrations. Experimental data indicate such low [Ca]/[Na] requires reaction temperatures in excess of 450 °C, notwithstanding high Fe solubility at such conditions. Using similar geochemical constraints to estimate maximum reaction temperatures (T_max) for the remainder of the dataset (as low as 80 °C, from off-axis fluids) yields a good (inverse) correlation between ⁸⁷Sr/⁸⁶Sr values and T_max. This trend is consistent with T vs. ⁸⁷Sr/⁸⁶Sr measured in < 200 °C hydrothermal carbonates from the Troodos Ophiolite [Weinzierl et al. (2018) Chem. Geol. in press]. Overall, these observations suggest that [Ca], [Na], [Sr] concentrations in hydrothermal fluids, and the associated ⁸⁷Sr/⁸⁶Sr, might be predictable based on precursor seawater chemistry and expected T_max over a more diverse range of oceanic hydrothermal settings than previously thought. This may also help constrain feedbacks between hydrothermal circulation and the variability of seawater chemistry over geologic timescales.