[S-GC16] Volatile Cycles in the Deep Earth - from Subduction Zones to the Mantle and Core

[SGC16-P01] Accurate measurement of H₂O concentration and speciation in silicate glasses using FTIR spectroscopy

*Iona McIntosh¹、Alexander Nichols¹、Kenichiro Tani²、Ed Llewellin³、Ian Schipper⁴、Bob Stewart⁵ (1.Japan Agency for Marine-Earth Science and Technology, Japan、2.National Museum of Nature and Science, Japan、3.Durham University, UK、4.Victoria University of Wellington, NZ、5.Massey University, NZ)

キーワード：FTIR spectroscopy, H2O speciation, volcanic glass

Accurate measurement of H₂O concentration and speciation in silicate glasses is important not only for studies of erupted glasses that use the volatile record to reconstruct eruption processes, but also for studies that use experimental glasses to find e.g. the partition coefficients of other volatiles that partition into H₂O-rich vapour. Fourier Transform Infrared (FTIR) spectroscopy can be used to find not only the overall H₂O concentration (H₂O_t) of silicate glasses, but also the individual concentrations of the two H₂O species: molecular H₂O (H₂O_m), and dissociated hydroxyl groups (OH). Here we discuss key developments and refinements of FTIR methodology with respect to finding the water contents of silicate glasses. Firstly, we demonstrate the importance of using a species-dependent H₂O_t molar absorptivity coefficient when using the 3500 cm^-1 H₂O_t absorbance peak to find H₂O_t and OH (indirectly as [OH] = [H₂O_t] – [H₂O_m]) concentrations, and in particular how this can be used to reconstruct the original H₂O contents of glasses that have since undergone secondary hydration at low temperature. Secondly, we discuss the strong dependence on glass composition of ‘silicate peaks’ at ~1830 cm^-1 and ~1600 cm^-1, and demonstrate how the 1830 cm^-1 peak can be used to correct FTIR imaging of H₂O concentrations in samples of varying glass thickness, and how the superposition of the 1630 cm^-1 H₂O_m and 1600 cm^-1 silicate peaks may result in overestimation of H₂O_m concentration in thick and/or H₂O_m-poor samples.