ABSTRACT

Analyses of elemental abundances by secondary ion mass spectrometry (SIMS) require matrix-matched standard samples to account for the matrix effect on correction factors. This requirement makes it difficult to obtain accurate results for geological samples of variable chemistry. In this study, we prepared 39 volcanic glasses of foiditic, basaltic, basaltic andesitic, rhyolitic and pure SiO₂ compositions, including synthetic samples and natural samples collected from the deep seafloor. The measured H₂O contents of these samples were in the range 0.02–4.8 wt%. We showed that calibration curves (H₂O content vs. ¹⁶OH^–/³⁰Si^–_SIMS ratio) differed according to the composition of the volcanic glasses. Our results demonstrated that for a particular ¹⁶OH/³⁰Si_SIMS ratio, water content could differ by up to a factor of five, depending on the composition of the volcanic glass. Although the correction factor (the slope of the calibration curve for water [H₂O/(¹⁶OH^–/³⁰Si^–)_SIMS]) was weakly correlated with SiO₂ content, we identified a stronger correlation with the molar weight (g mol^–1, on a one-oxygen mole basis) of the silicate glasses. Our results suggest that modification of the correction factor for the matrix effect on SIMS-based H₂O content of volcanic glasses according to their molar weights provides more accurate water contents of silicate glasses, regardless of their chemical composition and water content and without the need for a series of standard glasses of known water contents.