The effect of redox on the relative solubilities of copper and iron in Cl-bearing aqueous fluids at elevated temperatures and pressures: An experimental study with application to subseafloor hydrothermal systems

A series of hydrothermal experiments were performed to evaluate the role of redox, temperature, and fluid chemistry on the solubility of Cu- and Fe-bearing sulfide minerals in Na-K-Cl aqueous fluids. The experiments made use of potassic aluminosilicate minerals + quartz to maintain fluid pH at values generally appropriate for subseafloor hydrothermal systems. To avoid uncertainties in pH resulting from non-ideal activity-composition relations involving potassium aluminosilicate phases in aqueous fluids dominated by NaCl, however, fluid pH values were explicitly determined from magnetite solubility constraints. Total dissolved Cl concentrations of 0.3 to 1.94 molal were utilized, while redox constraints were established by different assemblages of oxide and sulfide phases. Relatively oxidizing and low dissolved S concentrations were maintained by the assemblage hematite-magnetite-pyrite (HMP), while the assemblage pyrite-pyrrhotite-magnetite (PPM) was used to maintain relatively reducing conditions and high dissolved S concentrations. Chalcopyrite was utilized in all experiments as a source of dissolved Cu. The use of flexible cell hydrothermal apparatus permitted sampling of fluids at experimental conditions. Phase equilibria constraints at 400°C, 500 bars indicate pH values between 4.8 and 5.5 for all dissolved Cl concentrations investigated. At a pH of 4.8, and dissolved NaCl equivalent to that of seawater (0.55 m), dissolved Cu decreased, and Fe increased with decreasing f_o2. For f_o2 values fixed by Hm-Mgt-Py and Py-Po-Mgt, dissolved Cu and Fe concentrations ranged from approximately 0.12 to 0.032 and 5.01 to 9.52 mmolal, respectively. As a consequence of this, Fe Cu ratio increased from approximately 50 to 250. Increasing pH from 4.8 to 5.0 affected significantly absolute metal concentrations, but this had only a slight affect on Fe Cu ratio. Dissolved concentrations of Cu and Fe were also affected by changes in total dissolved Cl and temperature, but again, Fe Cu ratio changed only slightly. The dominating effect of redox on Fe Cu ratio suggests that this parameter may be usefully applied to constrain redox in subseafloor hydrothermal systems. Fe Cu ratios of hot spring fluids at mid-ocean ridges and in back-arc basin settings reveal a relatively narrow range of values which are consistent with hydrothermal alteration under moderately oxidizing conditions. Fe Cu ratios measured for EPR, 21°N vent fluids approach values of 50, while Lau Basin vent fluids reveal values as high as 77. It is suggested here that anhydrite formation in subseafloor reaction zones helps to maintain relatively oxidizing conditions for these hot spring systems. In contrast, hot spring fluids from axial locations on the Mid-Atlantic Ridge (MARK) reveal Fe Cu ratios which indicate redox reactions buffered by the assemblage Py-Po-Mgt. Alternatively, hot spring fluids at MARK may have experienced extreme heat-loss effects with attendant loss of Cu very near the seafloor. Data from the present investigation and results of earlier basalt-fluid alteration experiments have shown that dissolved Cu is particularly sensitive to temperature at temperatures lower than approximately 350°C, and is lost from the fluid relative to Fe.