5:15 PM - 6:45 PM
[SGC32-P01] CO2 concentration of melt inclusions in drifted pumice from the 2021 Fukutoku-Oka-no-Ba eruption: Contribution of CO2 to explosive eruptions
Keywords:Fukutoku-Oka-no-Ba, Pumice, Explosive eruption, Carbon dioxide, Melt inclusion
Introduction: Factors that trigger explosive eruptions include, for example, fast magma ascent and low degassing rates within the conduit (e.g., Zhang, 1999), magma-water or ice interactions (e.g., Swanson et al., 2012), increased magma viscosity due to increased crystal content (e.g., Lavallée et al., 2007), and high volatile content (e.g., Sable et al., 2006). In particular, because CO2 has low solubility in magma, its concentration is one of the most important factors controlling magma explosion dynamics (Le Gall and Pichavant, 2016). In this study, we investigate the concentration of CO2 in olivine, clinopyroxene, and plagioclase hosted melt inclusions in drifted pumice to gain insight into the role of CO2 for the 2021 Fukutoku-Oka-no-Ba eruption (e.g., Yoshida et al., 2023, 2022).
Samples: Three pumices with different appearances were selected from the viewpoint of color. The pumices were amber, gray, and black in color and were collected from Amami-Oshima, Kikaijima, and Okinawa main islands, respectively. The phenocrysts were separated from the matrix using a high-voltage pulse selective crusher (SELLFRAG). Phenocrysts were hand-picked from the crushed samples and then polished to a mirror finish. The olivine hosted melt inclusions in the gray pumice tended to be homogeneous, whereas the other melt inclusions were often heterogeneous due to post-entrapment crystallization. To accurately estimate volatile concentrations from heterogeneous melt inclusions, these melt inclusions were heated to approximately 1300-1380°C in a heating stage, and quenched to homogenize them. The homogeneous melt inclusions consist of glass ± shrinkage bubble ± spinel.
Methods: CO2 in melt inclusions is contained in both glass and shrinkage bubbles, and in some cases, CO2 in shrinkage bubbles accounts for more than 90% of the total CO2 in the inclusions (Feignon et al., 2022). To estimate the CO2 concentration in melt inclusions, we measured the CO2 density of the shrinkage bubble in melt inclusions in ol, cpx, and pl using a Raman spectrometer (RAMAN touch VIS-HP-MAST; Nanophoton). A laser of 532 nm wavelength was focused on the sample through a 50x objective lens (N.A.=0.8, LUPlan; Nikon Corp.). The laser power at the sample surface was 21.7-74.8 mW, and the CO2 during analysis was above the critical temperature due to laser heating (Hagiwara et al. 2021b). The effects of underestimation of CO2 density due to laser heating and the influence of calibration curve discrepancies between laboratories were corrected (Hagiwara et al., 2021b, 2021a, 2020).
Results and Discussion: Despite the relatively high laser power, bright optical system, slow readout speed, and low number of readouts, CO2 was often below the detection limit in the Raman analysis. Even when detected, the CO2 intensity was weak and the analysis took a long time (6–27 minutes). Because the analyses were performed at relatively high laser power, we corrected for the effect of the apparent density reduction due to laser heating. However, because of the very low density, there was almost no difference in density before and after the correction (Hagiwara et al., 2021a). The CO2 density estimated from the Raman spectra of CO2 was almost constant regardless of the type of pumice or phenocryst, and no significant differences were observed (see figure). The average CO2 density in the shrinkage bubble measured in this study is about 0.014 g/cm3, which is one order of magnitude lower than that of other volcanic eruptions (e.g., 0.1–0.3 g/cm3; Sunset Crater (Arizona); Allison et al., 2021) in which CO2 is thought to have contributed significantly to explosive eruptions. Therefore, it is unlikely that the high CO2 concentration in the magma triggered the explosive eruption that occurred at Fukutoku-Oka-no-Ba, although only CO2 density data in the shrinkage bubble are currently available. This supports the conclusion of Yoshida et al. (2023) that nanolite crystallization triggered the eruption. In the future, to estimate the CO2 concentration in the inclusions more accurately, we will measure the inclusion volume by μ-X-ray CT and the CO2 concentration in the glass by SIMS. In addition, we will evaluate the effects of degassing and crystallization on the CO2 concentration based on the K/CO2 and Ba/CO2 ratios in order to investigate the cause of the relatively low CO2 concentration. For this purpose, the major and trace element compositions of melt inclusions will be measured.
Samples: Three pumices with different appearances were selected from the viewpoint of color. The pumices were amber, gray, and black in color and were collected from Amami-Oshima, Kikaijima, and Okinawa main islands, respectively. The phenocrysts were separated from the matrix using a high-voltage pulse selective crusher (SELLFRAG). Phenocrysts were hand-picked from the crushed samples and then polished to a mirror finish. The olivine hosted melt inclusions in the gray pumice tended to be homogeneous, whereas the other melt inclusions were often heterogeneous due to post-entrapment crystallization. To accurately estimate volatile concentrations from heterogeneous melt inclusions, these melt inclusions were heated to approximately 1300-1380°C in a heating stage, and quenched to homogenize them. The homogeneous melt inclusions consist of glass ± shrinkage bubble ± spinel.
Methods: CO2 in melt inclusions is contained in both glass and shrinkage bubbles, and in some cases, CO2 in shrinkage bubbles accounts for more than 90% of the total CO2 in the inclusions (Feignon et al., 2022). To estimate the CO2 concentration in melt inclusions, we measured the CO2 density of the shrinkage bubble in melt inclusions in ol, cpx, and pl using a Raman spectrometer (RAMAN touch VIS-HP-MAST; Nanophoton). A laser of 532 nm wavelength was focused on the sample through a 50x objective lens (N.A.=0.8, LUPlan; Nikon Corp.). The laser power at the sample surface was 21.7-74.8 mW, and the CO2 during analysis was above the critical temperature due to laser heating (Hagiwara et al. 2021b). The effects of underestimation of CO2 density due to laser heating and the influence of calibration curve discrepancies between laboratories were corrected (Hagiwara et al., 2021b, 2021a, 2020).
Results and Discussion: Despite the relatively high laser power, bright optical system, slow readout speed, and low number of readouts, CO2 was often below the detection limit in the Raman analysis. Even when detected, the CO2 intensity was weak and the analysis took a long time (6–27 minutes). Because the analyses were performed at relatively high laser power, we corrected for the effect of the apparent density reduction due to laser heating. However, because of the very low density, there was almost no difference in density before and after the correction (Hagiwara et al., 2021a). The CO2 density estimated from the Raman spectra of CO2 was almost constant regardless of the type of pumice or phenocryst, and no significant differences were observed (see figure). The average CO2 density in the shrinkage bubble measured in this study is about 0.014 g/cm3, which is one order of magnitude lower than that of other volcanic eruptions (e.g., 0.1–0.3 g/cm3; Sunset Crater (Arizona); Allison et al., 2021) in which CO2 is thought to have contributed significantly to explosive eruptions. Therefore, it is unlikely that the high CO2 concentration in the magma triggered the explosive eruption that occurred at Fukutoku-Oka-no-Ba, although only CO2 density data in the shrinkage bubble are currently available. This supports the conclusion of Yoshida et al. (2023) that nanolite crystallization triggered the eruption. In the future, to estimate the CO2 concentration in the inclusions more accurately, we will measure the inclusion volume by μ-X-ray CT and the CO2 concentration in the glass by SIMS. In addition, we will evaluate the effects of degassing and crystallization on the CO2 concentration based on the K/CO2 and Ba/CO2 ratios in order to investigate the cause of the relatively low CO2 concentration. For this purpose, the major and trace element compositions of melt inclusions will be measured.