TY - CONF

T1 - Volatile contents, degassing behaviour and hydration of early-erupted rhyolitic pyroclasts and ashes from Vulcan Chaitén, Chile

AU - Tuffen, Hugh

AU - Riley, Carina

AU - Castro, Jonathan M.

PY - 2010/12

Y1 - 2010/12

N2 - The concentration of dissolved volatile species in silicic pyroclasts, especially that of water, can yield key information about magma degassing and fragmentation depths. However, post-eruptive hydration can be highly problematic, with meteoric water overprinting magmatic water contents, especially in fine-grained or highly porous samples that have large surface area to volume ratios. In this study we have used thermogravimetric analysis-mass spectroscopy (TGA-MS) to characterize the volatile contents and degassing behaviour of a selection of juvenile pyroclasts and ashes erupted from Chaitén volcano, Chile in May-June 2008. The advantage of the selected samples is the short delay between eruption and collection (days to weeks), meaning that post-eruptive hydration ought to be minimal. The samples analysed, which were collected from proximal fallout deposits in June 2008, include a suite of centimetric pumices and obsidian clasts of varying porosity, the <250 μm ash fraction of the matrix, and a lapilli tuff. Samples were powdered to 125-500 μm and heated at 5 °C/min to 1250 °C after an initial 12 hour purge. Additionally, aliquots of ash samples were hydrated in distilled water for up to 3 months. The results indicate that pumices and obsidian samples have a similar total volatile content (0.7-1.6 wt %), which is dominated by water. The temperature of most rapid weight loss (dTGA peak) was significantly lower in the pumices (550-700 °C) than the dense obsidians (850-1000 °C), reflecting their greater surface area for diffusive degassing. Relationships between dTGA peak temperatures and total volatile contents are therefore different for pumices and obsidians, and results plot close to the non-hydrated obsidian endmember of Denton et al. (2009). Experimentally hydrated ash had two clearly separate dTGA peaks at 184-211 °C and 517-542 °C, which represent external and magmatic water respectively. The low-temperature peak is almost completely missing from natural samples, confirming that negligible post-eruptive hydration has occurred and indicating that the magmatic volatile content of even ashes and pumiceous clasts can be measured using bulk extraction techniques. Preliminary calculations of absolute water and CO2 concentrations from mass spectrometer data provide rough estimates of magma quenching depths to between 200-800 m beneath the surface, whilst pyroclast textures imaged in thin section and using scanning electron microscopy indicate heterogenous foam collapse to form dense pumices and obsidians. These results help to shed light on magma degassing during the ongoing Chaitén eruption and demonstrate that the bulk water content of freshly-collected pyroclasts may indeed be untainted by meteoric hydration. Reference Denton JS, Tuffen H, Gilbert JS, Odling N. (2009) The hydration and alteration of perlite and rhyolite from Iceland. Journal of the Geological Society of London 166, 895-904.

AB - The concentration of dissolved volatile species in silicic pyroclasts, especially that of water, can yield key information about magma degassing and fragmentation depths. However, post-eruptive hydration can be highly problematic, with meteoric water overprinting magmatic water contents, especially in fine-grained or highly porous samples that have large surface area to volume ratios. In this study we have used thermogravimetric analysis-mass spectroscopy (TGA-MS) to characterize the volatile contents and degassing behaviour of a selection of juvenile pyroclasts and ashes erupted from Chaitén volcano, Chile in May-June 2008. The advantage of the selected samples is the short delay between eruption and collection (days to weeks), meaning that post-eruptive hydration ought to be minimal. The samples analysed, which were collected from proximal fallout deposits in June 2008, include a suite of centimetric pumices and obsidian clasts of varying porosity, the <250 μm ash fraction of the matrix, and a lapilli tuff. Samples were powdered to 125-500 μm and heated at 5 °C/min to 1250 °C after an initial 12 hour purge. Additionally, aliquots of ash samples were hydrated in distilled water for up to 3 months. The results indicate that pumices and obsidian samples have a similar total volatile content (0.7-1.6 wt %), which is dominated by water. The temperature of most rapid weight loss (dTGA peak) was significantly lower in the pumices (550-700 °C) than the dense obsidians (850-1000 °C), reflecting their greater surface area for diffusive degassing. Relationships between dTGA peak temperatures and total volatile contents are therefore different for pumices and obsidians, and results plot close to the non-hydrated obsidian endmember of Denton et al. (2009). Experimentally hydrated ash had two clearly separate dTGA peaks at 184-211 °C and 517-542 °C, which represent external and magmatic water respectively. The low-temperature peak is almost completely missing from natural samples, confirming that negligible post-eruptive hydration has occurred and indicating that the magmatic volatile content of even ashes and pumiceous clasts can be measured using bulk extraction techniques. Preliminary calculations of absolute water and CO2 concentrations from mass spectrometer data provide rough estimates of magma quenching depths to between 200-800 m beneath the surface, whilst pyroclast textures imaged in thin section and using scanning electron microscopy indicate heterogenous foam collapse to form dense pumices and obsidians. These results help to shed light on magma degassing during the ongoing Chaitén eruption and demonstrate that the bulk water content of freshly-collected pyroclasts may indeed be untainted by meteoric hydration. Reference Denton JS, Tuffen H, Gilbert JS, Odling N. (2009) The hydration and alteration of perlite and rhyolite from Iceland. Journal of the Geological Society of London 166, 895-904.

M3 - Poster

T2 - AGU Fall Meeting 2010

Y2 - 13 December 2010 through 17 December 2010

ER -