Dehydration and deprotonation processes in minerals: development of new spectroscopic techniques

Abstract

A convenient way to address in detail the mechanisms of dehydration and deprotonation processes in minerals is by using in situ-FTIR spectroscopy. This technique allows measuring the variations in the arrangements of either OH or H2O molecular groups at the ppm scale, in real-time and directly in operando. In this work, high temperature (HT) FTIR methods were preliminarily set-up by studying the dehydration of the Fe-sulfate hohmannite, and then extensively used to study the deprotonation process of iron-rich amphiboles. The thermal transformation of hohmannite, Fe2[O(SO4)2] 8H2O, a secondary iron-bearing hydrous sulfate, was investigated by in situ HT-FTIR spectroscopy. Spectroscopic data were used to complement the X-ray powder diffraction results, allowing to better define the reaction paths for homannite and the stability of its HT products. Five dehydration/transformation steps were identified in the heating range of 25 – 800 °C (Ventruti et al., 2015). Previous HT studies on amphiboles (e.g. Addison et al., 1962a; Ernst and Wai, 1970; Hodgson et al., 1965; Ungaretti, 1980), pointed out a significant coupling between deprotonation and oxidation of Fe2+. More recently, oxidation reactions in Fe-rich amphiboles have been considered to be responsible for the enhanced electrical conductivity in subduction zones (Wang et al., 2012). Detailed HT-FTIR experiments were done on a synthetic Fe-richterite close to Na(Na,Ca)2Fe2+ 5Si8O22(OH)2 and on a riebeckite from Malawi, close to Na2Fe3+ 2 Fe2+ 3Si8O22(OH)2. A multi-methodological approach was used, including HT-FTIR spectroscopy, X-ray powder diffraction, single crystal structural refinement (SREF), and Mössbauer spectroscopy, to characterize possible structural changes induced by increasing temperature. The main results obtained during this study are: Abstract 1) IR spectra collected in situ for both amphiboles show an anomalous and strong increase in the OH-stretching absorption as a function of T, whereas data collected on quenched samples show no OH-loss before the deprotonation temperature. This feature has never been observed before for OH-bearing minerals and cannot be explained based on the tilting of the O-H vector with T as proposed, for example, for micas (Tokiwai and Nakashima, 2010). We have no final explanation at present for this issue, but such an increase of the OH-stretching intensity seems to be related to the OH-loss mechanism which is active in the studied amphiboles. 2) X-ray single crystal structural refinements (XR-SREF) done by Oberti et al. (2016) on potassic-ferro-richterite and those done here on riebeckite are consistent with an oxidation of Fe2+ to Fe3+ occurring completely at M(1). The same result was obtained by Mössbauer spectroscopy done on heat-treated riebeckite. Refinement of cell-parameters for both potassic-ferro-richterite (Oberti et al., 2016) and riebeckite (this study) showed an abrupt contraction of the unit cell dimensions in the 350- 450°C, and 400-500°C T range for potassic-ferro-richterite and riebeckite respectively. Such a feature is compatible with oxidation of Fe2+ at the octahedral sites, coupled to H-loss, responsible in particular, for the contraction along the a* crystallographic direction. HT-FTIR experiments done on both amphiboles show in fact the disappearance of the OH-stretching absorption in the same T range, although a significant shift is observed in the deprotonation T when comparing data from both techniques. This shift is particularly significant for data collected on single- crystals. 3) HT-FTIR experiments done on riebeckite powders embedded in KBr matrix showed no hydrogen loss after 180 minutes at 520°C, suggesting that, in agreement with previous authors (Addison et al., 1962a; Ernst and Wai, 1970) the Fe-oxidation process, coupled/responsible for the deprotonation process, occurs basically at the crystal surface, where interaction with the atmospheric oxygen is possible. HT-FTIR experiments done on single crystals using a N2 flux are compatible with this scenario. On the basis of these results, the deprotonation of Fe-amphiboles can be schematically represented as: 4Fe2+ + 4OH- + O2 = 4Fe3+ + 4O2- + 2H2O Abstract The reactants Fe2+ and OH- are provided by the amphibole, while O2 is provided by the external atmosphere. Single-crystal XRD demonstrates that Fe3+ and O2 2- on the right-side of this equation are still part of the (oxo)-amphibole, while only H+ is released, by reacting at the crystal surface with external O2 to produce H2O. For the reaction to proceed, there is the need of continuous availability of Fe2+ and H at the crystal surface, and these must be provided by inside the crystal, via electron and proton diffusion. Comparison on the XRD and spectroscopic data suggests that, as expected, the electron diffusion is faster that the H diffusion. Isothermal experiments done on powdered riebeckite yielded an activation energy (Ea) for the H diffusion of 20 ± 2 kJ/mol. Key words: deprotonation of amphiboles, potassic-ferro-richterite, riebeckite, HT- FTIR spectroscopy, single-crystal and powder HT X-ray diffraction, EMP analysis, Mössbauer spectroscopy.