Physics

Geology

Organic chemistry

Andesite

Molar absorptivity

Absorbance

Environmental chemistry

Analytical Chemistry (journal)

Optics

Volcano

Infrared spectroscopy

Composite material

Mineralogy

Seismology

Albite

Chemistry

Volcanic rock

Materials science

Silicate

Quartz

Silicate glass

Chromatography

American Mineralogist

Fourier transform infrared (FTIR) spectroscopy can be used to determine the concentration and speciation of dissolved water in silicate glasses if the molar absorptivity coefficients (ε) are known. Samples that are thin and/or water-poor typically require the use of the mid-IR 3500 cm−1 total water (H2Ot) and 1630 cm−1 molecular water (H2Om) absorbance bands, from which hydroxyl water (OH) must be determined by difference; however, accurate determination of H2Ot and OH is complicated because ε3500 varies with water speciation, which is not usually known a priori. We derive an equation that uses end-member ε3500 values to find accurate H2Ot and OH concentrations from the 3500 cm−1 absorbance for samples where only the H2Om concentration need be known (e.g., from the 1630 cm−1 band). We validate this new species-dependent ε3500 method against published data sets and new analyses of glass standards. We use published data to calculate new end-member ε3500 values of ε3500OH = 79 ± 11 and ε3500H2Om = 49 ± 6 L/mol·cm for Fe-bearing andesite and ε3500OH = 76 ± 12 and ε3500H2Om = 62 ± 7 L/mol·cm for Fe-free andesite. These supplement existing end-member values for rhyolite and albite compositions. We demonstrate that accounting for the species-dependence of ε3500 is especially important for hydrated samples, which contain excess H2Om, and that accurate measurement of OH concentration, in conjunction with published speciation models, enables reconstruction of original pre-hydration water contents. Although previous studies of hydrous silicate glasses have suggested that values of ε decrease with decreasing tetrahedral cation fraction of the glass, this trend is not seen in the four sets of end-member ε3500 values presented here. We expect that future FTIR studies that derive end-member ε3500 values for additional compositions will therefore not only enable the species-dependent ε3500 method to be applied more widely, but will also offer additional insights into the relationship between values of ε and glass composition.

Accounting for the species-dependence of the 3500 cm−1H2Otinfrared molar absorptivity coefficient: Implications for hydrated volcanic glasses