An experimental study of hydroxyl in quartz using infrared spectroscopy and ion microprobe techniques

Abstract

We have measured the concentrations of hydroxyl, deuterium, Al, Fe, Li, Na, K, and Rb in a natural quartz crystal before and after hydrothermal treatment at 1.5 GPa and 800°–1050°C. We employed microbeam infrared spectroscopy and ion probe techniques to avoid impurities trapped in healed cracks and fluid inclusions that might bias a normal bulk analysis. The ƒH2 of our experiments were buffered to the hematite‐magnetite‐(OH) fluid, nickel‐nickel oxide‐(OH) fluid, or iron‐wustite‐(OH) fluid phase assemblages. After hydrothermal treatment, the samples contained local concentrations of hydrogen or deuterium of several hundred atoms/106 Si (the starting crystal contained 45 H/106 Si). We did several experiments with Al2O3 or RbCl added to the sample charge and found local Al enrichment where the deuterium concentration was high but no Rb enrichment. Finally, we measured trace elements and hydroxyl in a quartz sample after plastic deformation in a talc furnace assembly; in regions of the sample containing basal and prismatic deformation lamellae (but no visible healed microcracks at 400× optical magnification) hydroxyl had increased to ∼200 OH/106 Si with no increase in Al or Fe. Samples enriched in hydroxyl but not Al (including the plastically strained sample) gave infrared spectra resembling natural amethyst crystals. We observed that the sharp pleochroic peaks near ∼3400 cm−1 and present in the starting crystal, were very intense only in samples showing Al enrichment, whereas the intensity of the sharp pleochroic peaks near 3600 cm−1 and broad isotropic absorption were independent of Al. Our analyses indicate that more hydrogen was introduced into the treated samples than Al or Fe. Because one proton or alkali cation is needed to screen each Al or Fe atom substituted into a Si lattice site, we conclude that the hydrothermal treatment had produced new hydroxyl defects in the quartz that did not contain Al or Fe. Although the speciation of this excess hydroxyl is unknown, it is present in all varieties of quartz that show hydrolytic weakening: synthetic quartz, amethyst, hydrothermally treated natural quartz crystals, and natural quartz deformed in talc assemblies. In the absence of microcracking or solution‐precipitation mechanisms that may mechanically trap OH or H2O molecules, we suggest that hydrogen diffusion, and reaction with lattice oxygen, may introduce hydroxyl defects into quartz and contribute to hydrolytic weakening.