Vesiculation in rhyolite at low H2O contents: A thermodynamic model

• Published in: Geochemistry, Geophysics, Geosystems
• Main Authors: Ryan, Amy G., Russell, James K., Hess, Kai-Uwe, Phillion, A.B., Dingwell, Donald B.
• Other Authors: Materials Science and Engineering
• Language: English
• Published: Wiley 2015
• Subjects: Krafla; Hrafntinnuhryggur; Iceland
• Online Access: http://hdl.handle.net/11375/21523; https://doi.org/10.1002/2015GC006024

We present experimental data on the thermodynamics and kinetics of bubble nucleation and growth in weakly H2O-oversaturated rhyolitic melts. The high-temperature (900-1100°C) experiments involve heating of rhyolitic obsidian from Hrafntinnuhryggur, Krafla, Iceland to above their glass transition temperature (Tg ~ 690°C) at 0.1 MPa for times of 0.25-24 hours. During experiments, the rhyolite cores increase in volume as H2O vapour-filled bubbles nucleate and expand. The extent of vesiculation, as tracked by porosity, is mapped in temperature-time (T-t) space. At constant temperature and for a characteristic dwell time, the rhyolite cores achieve a maximum volume where the T-t conditions reach thermochemical equilibrium. For each T-t snapshot of vesiculation, we use 3D analysis of X-ray computed tomographic (XCT) images of the quenched cores to obtain the bubble number density (BND) and bubble size distribution (BSD). BNDs for the experimental cores are insensitive to T and t, indicating a single nucleation event. All BSDs converge to a common distribution, independent of T, melt viscosity (η), or initial degree of saturation, suggesting a common growth process. We use these data to calibrate an empirical model for predicting the rates and amounts of vesiculation in rhyolitic melts as a function of η and thermochemical affinity (A): two computable parameters that are dependent on T, pressure and H2O content. The model reproduces the experimental dataset and data from the literature to within experimental error, and has application to natural volcanic systems where bubble formation and growth are not diffusion limited (e.g., lavas, domes, ignimbrites, conduit infill). Canadian Foundation for Innovation