UV–Vis spectrophotometric and XAFS studies of ferric chloride complexes in hyper-saline LiCl solutions at 25–90 °C

The speciation and thermodynamic properties of ferric chloride complexes in hydrothermal solutions and hypersaline brines are still poorly understood, despite the importance of this element as a micronutrient and ore-component. Available experimental data are limited to room temperature and relatively low chloride concentrations. This paper reports results of UV–Vis spectrophotometric and synchrotron XAFS experiments of ferric chloride complexes in chloride concentrations up to 15 m and at temperatures of 25–90 °C. Qualitative interpretation of the UV–Vis spectra shows that FeCl2+, FeCl2+, FeCl3(aq) and FeCl4− were present in the experimental solutions. As chloride concentrations increase, higher ligand number complexes become important with FeCl4− predominating in solutions containing more than 10 m at 25 °C. The predominance fields of FeCl3(aq) and FeCl4− expand to lower Cl concentrations with increasing T. Both XANES and UV–Vis spectra reveal a major change in the geometry of the complex between FeCl2+ and FeCl3(aq). EXAFS data confirm that the number of chloride ligands increases with increasing chloride concentration and show that Fe3+, FeCl2+ and FeCl2+ share an octahedral geometry. FeCl3(aq) could be either tetrahedral or trigonal dipyramidal, while FeCl4− is expected to be tetrahedral. EXAFS data support a tetrahedral geometry for FeCl4−, especially at 90 °C, but do not allow to distinguish between a tetrahedral or trigonal dipyramidal geometry for FeCl3(aq) because of similar Fe–Cl distances. At room temperature, EXAFS data suggest that FeCl3(aq) may be a mixture of octahedral and tetrahedral or trigonal dipyramidal forms. om temperature formation constants for three ferric chloride complexes (FeCl2+, FeCl3(aq) and FeCl4−) determined from the UV data are generally in good agreement with previous studies. Calculations based on the properties extrapolated to 300 °C show that hematite solubility is much higher than previously estimated, and that the high orders complexes FeCl3(aq) and FeCl4− are important at high temperatures even in solutions with low chloride concentrations. The accuracy of these properties is limited by a poor understanding of activity–composition relationships in concentrated electrolytes, and by limitations in the available experimental techniques and extrapolation algorithms; however, the inclusion of higher order complexes in numerical models of ore transport and deposition allows for a more accurate qualitative prediction of Fe behaviour in hydrothermal and hypersaline systems.