TY - JOUR

T1 - CO2 quantification in silicate glasses using µ-ATR FTIR spectroscopy

AU - Schanofski, Maximilian

AU - Koch, Lennart

AU - Schmidt, Burkhard C.

N1 - Publisher Copyright: © 2023 by Mineralogical Society of America.

PY - 2023/7/3

Y1 - 2023/7/3

N2 - A new method for measurements of high-CO2 concentrations in silicate glasses was established using micro-attenuated total reflectance (μ-ATR) Fourier transform infrared (FTIR) spectroscopy in the mid-IR (MIR) region. We studied two glass/melt compositions, namely leucitite and granite, to cover samples in which CO2 is dissolved as carbonate ions C O 3 2 - $\left(\mathrm{CO}_{3}^{2-}\right)$or as CO2 molecules C O 2 m o l. $\left(\mathrm{CO}_{2}^{\mathrm{mol}}\right).$In the 3 leucitite glasses a carbonate absorption doublet with maxima at 1510 and 1430 cm-1 has shown to clearly separate from aluminosilicate lattice vibrations at lower wavenumbers. Due to the lower sensitivity of the μ-ATR method, we were able to measure high-CO2 contents c C O 2 > 0.5 w t % $\left(\mathrm{c}_{\mathrm{CO}_{2}}>0.5 \mathrm{wt} \text{%}\right)$in experimental silicate glasses that would only be measurable with great dificulties using established transmission MIR measurements due to detector linearity limit efects even with very thin sample wafers. The peak heights of the 1430 cm-1 ATR band (A1430), normalized to the integral of the T-O lattice vibrations (T = Si, Al, Fe) at about 930 cm-1 (Int930) show a linear trend with CO2 contents in the range 0.2-4.3 wt%, yielding a linear correlation with cCO2 (wt%) = 0.4394 ± 0.006·A1430·10000/Int930. The normalization of the CO2 related band to a lattice vibration accounts for variations in the quality of contact between ATR crystal and sample, which has a direct effect on signal intensity. In granitic glasses, where CO2 is dissolved as C O 2 m o l $\mathrm{CO}_{2}^{\mathrm{mol}}$only, the asymmetric stretching vibration at 2350 cm-1 overlaps with the signal of atmospheric, gaseous CO2. As the ATR signal of dissolved CO2 is very weak, the atmospheric signal may dominate the spectrum. Since the absorbance spectrum is calculated by division of the single-channel sample spectrum by a single-channel reference spectrum measured in air, keeping the laboratory and spectrometer atmosphere as constant as possible during spectral acquisition can resolve the problem. Nonetheless, a procedure to subtract the signal of remaining atmospheric CO2 may still be required for the spectral evaluation. We studied a series of 5 granitic glasses with C O 2 m o l $\mathrm{CO}_{2}^{\mathrm{mol}}$contents of 0.08 to 0.27 wt% and found an excellent linear relation between CO2 concentration and lattice vibration normalized ATR intensity of the 2350 cm-1 band: c C O 2 (w t %) = $\mathrm{c}_{\mathrm{CO}_{2}}(\mathrm{wt} \text{%})=$0.2632 ± 0.0016·A2350·10000/Int990. Although the C O 2 m o l $\mathrm{CO}_{2}^{\mathrm{mol}}$concentrations in our granitic glass series can still be analyzed without major difficulties by conventional transmission IR spectroscopy, our data demonstrate the potential of the ATR method for samples with higher CO2 contents or for samples where a high spatial resolution is required (melt inclusions, vesicular or partially crystallized glasses). The lower limits of the ATR method are approximately 0.2 wt% CO2 dissolved as carbonate groups or 0.1 wt% CO2 (or slightly less) dissolved in molecular form.

AB - A new method for measurements of high-CO2 concentrations in silicate glasses was established using micro-attenuated total reflectance (μ-ATR) Fourier transform infrared (FTIR) spectroscopy in the mid-IR (MIR) region. We studied two glass/melt compositions, namely leucitite and granite, to cover samples in which CO2 is dissolved as carbonate ions C O 3 2 - $\left(\mathrm{CO}_{3}^{2-}\right)$or as CO2 molecules C O 2 m o l. $\left(\mathrm{CO}_{2}^{\mathrm{mol}}\right).$In the 3 leucitite glasses a carbonate absorption doublet with maxima at 1510 and 1430 cm-1 has shown to clearly separate from aluminosilicate lattice vibrations at lower wavenumbers. Due to the lower sensitivity of the μ-ATR method, we were able to measure high-CO2 contents c C O 2 > 0.5 w t % $\left(\mathrm{c}_{\mathrm{CO}_{2}}>0.5 \mathrm{wt} \text{%}\right)$in experimental silicate glasses that would only be measurable with great dificulties using established transmission MIR measurements due to detector linearity limit efects even with very thin sample wafers. The peak heights of the 1430 cm-1 ATR band (A1430), normalized to the integral of the T-O lattice vibrations (T = Si, Al, Fe) at about 930 cm-1 (Int930) show a linear trend with CO2 contents in the range 0.2-4.3 wt%, yielding a linear correlation with cCO2 (wt%) = 0.4394 ± 0.006·A1430·10000/Int930. The normalization of the CO2 related band to a lattice vibration accounts for variations in the quality of contact between ATR crystal and sample, which has a direct effect on signal intensity. In granitic glasses, where CO2 is dissolved as C O 2 m o l $\mathrm{CO}_{2}^{\mathrm{mol}}$only, the asymmetric stretching vibration at 2350 cm-1 overlaps with the signal of atmospheric, gaseous CO2. As the ATR signal of dissolved CO2 is very weak, the atmospheric signal may dominate the spectrum. Since the absorbance spectrum is calculated by division of the single-channel sample spectrum by a single-channel reference spectrum measured in air, keeping the laboratory and spectrometer atmosphere as constant as possible during spectral acquisition can resolve the problem. Nonetheless, a procedure to subtract the signal of remaining atmospheric CO2 may still be required for the spectral evaluation. We studied a series of 5 granitic glasses with C O 2 m o l $\mathrm{CO}_{2}^{\mathrm{mol}}$contents of 0.08 to 0.27 wt% and found an excellent linear relation between CO2 concentration and lattice vibration normalized ATR intensity of the 2350 cm-1 band: c C O 2 (w t %) = $\mathrm{c}_{\mathrm{CO}_{2}}(\mathrm{wt} \text{%})=$0.2632 ± 0.0016·A2350·10000/Int990. Although the C O 2 m o l $\mathrm{CO}_{2}^{\mathrm{mol}}$concentrations in our granitic glass series can still be analyzed without major difficulties by conventional transmission IR spectroscopy, our data demonstrate the potential of the ATR method for samples with higher CO2 contents or for samples where a high spatial resolution is required (melt inclusions, vesicular or partially crystallized glasses). The lower limits of the ATR method are approximately 0.2 wt% CO2 dissolved as carbonate groups or 0.1 wt% CO2 (or slightly less) dissolved in molecular form.

KW - ATR FTIR

KW - ATR-micro spectroscopy

KW - carbon dioxide

KW - CO

KW - CO quantification

KW - silicate glasses

UR - http://www.scopus.com/inward/record.url?scp=85149433623&partnerID=8YFLogxK

U2 - 10.2138/am-2022-8477

DO - 10.2138/am-2022-8477

M3 - Article

VL - 108

SP - 1346

EP - 1356

JO - American mineralogist

JF - American mineralogist

SN - 0003-004X

IS - 7

ER -