ようこそ、山下茂です

Shigeru Yamashita's

Reseach Interests


Vibrational spectroscopy of water-bearing silicate glasses

Micro-FTIR spectroscopy is a powerful analytical method for water species dissolved in silicate glasses. One disadvantage of this method is that the molar absorptivities of water species have a composition dependence; the molar absorptivities of water species bands (2.8µm band = fundamental stretching of OH-group and molecular H2O, 2.2µm band = combination mode of fundamental stretching / bending of OH-group, 1.9µm band = combination mode of fundamental stretching / bending of molecular H2O, etc.) generally increas as silica content of glass increases (e.g., Yamashita et al., 1997). This, together with insufficient data coverage of glass composition space in previously published works, has prevented simple adaptation of this method for water analysis of a silicate glass of composition of interest. The aim of this research is to make a reconnaissance of the effect of glass composition on the molar absorptivities of water species, over the range of glass composition of common magmatic melts.

I calibrated the molar absorptivities of dissolved water species bands to a series of standard samples (Table; basalt and dacite glasses, Yamashita et al., 1997; rhyolite glasses, Yamashita, 1999; Yamashita, unpublished) that have been synthesized in an internally heated pressure vessel and a piston cylinder apparatus. Excellent agreement was found between the water contents calculated from the absorption peak heights and those determined by H2 manometry or calculated from the weight proportion of loaded water, when the Lambert-Beer's law and the following molar absorptivities (m3/mol m) were used for each glass composition:epsilon2.8µm = 6.4ア0.1(1s), epsilon2.2µm = 0.085˜ 0.007, and epsilon1.9µm = 0.084ア0.006 for tholeiite and high-Al basalt glasses; epsilon2.8µm = 6.8˜ 0.1, epsilon2.2µm = 0.094˜ 0.006, and epsilon1.9µm = 0.16˜ 0.03 for dacite glasses; epsilon2.8µm = 7.2˜ 0.4, epsilon2.2µm = 0.199˜ 0.009, and epsilon1.9µm = 0.158˜ 0.005 for rhyolite glasses (Figure).

This data coverage of composition space is practically sufficient, making it possible to determine water contents of glasses of common magmatic melt compositions, with a typical analytical precision of 5%. We have also developed a simple computer code for subtraction of the background spectrum, which ensures reproducibility of analytical results.

(Left) A micro-FTIR spectrometer (Micro-Janssen, Nihon Bunko Co.) installed in our lab. Calibration was performed with this spectrometer. This spectrometer has Cassegrain optics, a broad band MCT detector, a KBr beamsplitter, a nichrome wire mid-IR light source, and a W near-IR light source. Sample glasses were mounted in orthodontic resin and doubly polished. The IR beam was passed through a 100 µm x 100 µm sample spot by adjusting an aperture window.

(Below)Calibration of molar absorptivities (basalt and dacite glasses, Yamashita et al., 1997; rhyolite glasses, Yamashita, 1999; Yamashita, unpublished). Error bars represent propagation of the uncertainties in sample thickness, sample density, and absorption peak height determination.

Anhydrous compositions of samples (*Yamashita et al., 1997; **Yamashita, 1999)

(wt%) *Tholeiite basalt *High-Al basalt *Dacite **Rhyolite
SiO2 53.59 51.84 65.42 77.38
TiO2 1.23 1.39 0.86 0.17
Al2O3 14.05 17.08 15.26 12.33
FeO 12.59 10.20 6.08 1.26
MnO 0.22 0.18 0.15 0.05
MgO 5.26 5.32 1.57 0.17
CaO 10.41 10.00 5.73 1.11
Na2O 2.06 2.87 3.98 3.39
K2O 0.43 0.79 0.78 3.58
P2O5 0.16 0.33 0.17 0.01

References:


Link to: Shigeru Yamashita's Research Interests