- © The Mineralogical Society Of America
A large number of studies of water in pyroxenes have confirmed that essentially all pyroxenes of terrestrial origin as a rule contain substantial amounts of OH, varying as a function of chemical composition and geological occurrence. The highest OH-concentrations are observed in mantle-derived samples, and it is now well established that both clino- and orthopyroxene can be considered as major host minerals for water in the mantle (e.g., Bell and Rossman 1992; Bell et al. 2004; Bolfan-Casanova 2005).
The first firm observations of OH in pyroxenes were made more than three decades ago. Even earlier, the “excess” water recorded in many wet-chemical analyses of pyroxenes and other nominally anhydrous minerals had been taken as indications of structurally bonded water, although it was difficult to confirm whether this “excess” water was really due to intrinsic OH, or if it was caused by fluid or solid inclusions, or contaminants on cracks and surfaces. A water incorporation model based on OH ions replacing O at their normal structural positions in pyroxenes was proposed by Martin and Donnay (1972), who also suggested that OH-containing nominally anhydrous minerals could provide the main storage mechanism for water in the mantle. Somewhat later Wilkins and Sabine (1973) published IR spectra of several nominally anhydrous minerals, including a diopside from crustal occurrence (Fig. 1⇓). The spectra contained sharp OH absorption bands which were verified by deuterium exchange, and the water contents were also analyzed by an electrolytic technique, which showed considerably lower concentrations than what had earlier been indicated from wet-chemical analyses. Also Runciman et al. (1973) observed absorption bands in the OH range in a spectroscopic study in the full infrared to UV-range of an enstatite from Kimberley, South Africa, which they suspected to be caused by traces of OH. A more …