TY - JOUR

T1 - Water release from dacitic melts near the glass transition and implications for volcanic eruptions

AU - Pohl, Florian

AU - Behrens, Harald

AU - Dohmen, Ralf

N1 - Funding Information: The research was supported by the German Science foundation (project BE 1720/43-1 and researcher unit FOR2881). The authors thank Julian Feige for preparation of thin sections, as well as Viona Klamt for her assistance with infrared and Raman measurements. Youxue Zhang and Kai-Uwe Hess helped to improve the quality of the paper by fruitful comments.

PY - 2024/2/5

Y1 - 2024/2/5

N2 - To get insights in the mechanisms of magma degassing and fragmentation, we performed dehydration experiments at ambient pressure with dacitic glasses containing between 1.42 and 5.28 wt% H2O. The focus was on the degassing behavior in the range of glass transition, i.e., from temperatures 50 K below to 50 K above the glass transition temperature Tg. At each T the duration of the experiments was varied between 5 and 125 h. Water contents and initial water speciation was determined by FTIR spectroscopy. Diffusion depth profiles were measured by Raman spectroscopy. We developed an equation to calculate the total water content along the depth profiles from the ratio of the fundamental OH stretching vibration band at 3550 cm−1 and the T-O-T bending band near 500 cm−1. Glasses with high water content (4.66–5.28 wt%) showed huge alterations, i.e., crack formation below Tg and crystallization at and above Tg, but glasses with water contents up to 2.5 wt% were visually unchanged and suitable for water diffusion analysis. The measured profiles of total water content vs. distance can be fitted well by an error function, assuming that the water content in the glass does not decrease to zero at the surface. However, the apparent water concentration at the surface and the total water diffusivity derived from the error function fitting decrease with increasing duration of the experiment. This clearly shows that degassing of dacitic melts is not a simple out-diffusion mechanism. The model of Coumans et al. (2020) that explicitly considers the kinetics of the interconversion of water species can reproduce the temporal evolution of the total water concentration profiles very well. Our results show the importance of considering explicitly the interconversion kinetics of water species when modelling water diffusion in melts at high viscosities. Due to the slow interconversion of OH groups to much more mobile water molecules, dehydration is much slower in the range of glass transition than predicted by diffusion models based on experiments performed on stable melts at magmatic temperatures.

AB - To get insights in the mechanisms of magma degassing and fragmentation, we performed dehydration experiments at ambient pressure with dacitic glasses containing between 1.42 and 5.28 wt% H2O. The focus was on the degassing behavior in the range of glass transition, i.e., from temperatures 50 K below to 50 K above the glass transition temperature Tg. At each T the duration of the experiments was varied between 5 and 125 h. Water contents and initial water speciation was determined by FTIR spectroscopy. Diffusion depth profiles were measured by Raman spectroscopy. We developed an equation to calculate the total water content along the depth profiles from the ratio of the fundamental OH stretching vibration band at 3550 cm−1 and the T-O-T bending band near 500 cm−1. Glasses with high water content (4.66–5.28 wt%) showed huge alterations, i.e., crack formation below Tg and crystallization at and above Tg, but glasses with water contents up to 2.5 wt% were visually unchanged and suitable for water diffusion analysis. The measured profiles of total water content vs. distance can be fitted well by an error function, assuming that the water content in the glass does not decrease to zero at the surface. However, the apparent water concentration at the surface and the total water diffusivity derived from the error function fitting decrease with increasing duration of the experiment. This clearly shows that degassing of dacitic melts is not a simple out-diffusion mechanism. The model of Coumans et al. (2020) that explicitly considers the kinetics of the interconversion of water species can reproduce the temporal evolution of the total water concentration profiles very well. Our results show the importance of considering explicitly the interconversion kinetics of water species when modelling water diffusion in melts at high viscosities. Due to the slow interconversion of OH groups to much more mobile water molecules, dehydration is much slower in the range of glass transition than predicted by diffusion models based on experiments performed on stable melts at magmatic temperatures.

KW - Dacitic glass

KW - Dehydration

KW - Raman spectroscopy

KW - Water diffusion

KW - Water speciation

UR - http://www.scopus.com/inward/record.url?scp=85180540432&partnerID=8YFLogxK

U2 - 10.1016/j.chemgeo.2023.121885

DO - 10.1016/j.chemgeo.2023.121885

M3 - Article

AN - SCOPUS:85180540432

VL - 645

JO - Chemical geology

JF - Chemical geology

SN - 0009-2541

M1 - 121885

ER -