Experimental investigation of ascending magmas
Zusammenfassung der Projektergebnisse
Magma degassing during ascent can be simulated by decompression experiments with volatile-bearing silicate melts. Decompression-induced volatile supersaturation in the melt results in bubble nucleation and growth. The simulation of these dynamic degassing processes necessitates optimized experiments in which the results are primarily dependent on the decompression rate. This project reveals the influence of the decompression methods (continuous/step-wise decompression, CD/SD) reported in literature on homogeneous bubble nucleation and the course of H2O degassing. H2O-bearing silicate melts were decompressed at a super-liquidus temperature of 1323 K in an internally heated argon pressure vessel. Decompression started from a pressure of 200 MPa with nominal decompression rates of 0.0028-1.7 MPa·s^-1. At final pressures of 100-75 MPa, the samples were quenched rapidly at isobaric conditions. The experiments document that melt degassing is highly sensitive to the experimental protocol. Based on the results of optimized experiments, this project provides essential guidelines for the conduction and interpretation of degassing experiments. A fundamentally important aspect is the consideration of the bubble volume reduction due to decreasing molar volume (Vm) of the exsolved H2O during isobaric rapid quench. The bubble volumes and the porosity of a vitrified sample do not correspond to the condition prior to cooling. The bubble volume can be corrected using a shrinking factor that is dependent on the Vm (H2O) at run temperature prior to isobaric rapid quench (Trq) and at a fictive temperature (Tf) where further shrinkage is prevented by melt viscosity. In first approximation, the glass transition temperature of the melt can be used for Tf. The corrected porosity of a quenched sample is within a relative error of <10 % equal to the expected porosity at Trq that can be calculated from the residual H2O content in the glass. To date, the volume reduction of bubbles during cooling under high pressure has not been considered in the analysis and interpretation of vesiculated samples. Diffusive bubble growth during decompression can result in volatile concentration gradients towards bubbles. The evaluation of micro-FTIR imaging with a focal plane array detector as analytical method to study diffusion processes of CO2 and H2O in silicate melts was also part of this project. Bulk CO2 and H2O diffusivities derived from imaging of concentration-distance profiles in diffusion couple samples are within the error identical to the diffusivities derived from standard single-element detector analysis. The imaging technique is particularly suited to study diffusion processes on a micrometer scale. However, imaging of the total H2O concentration around a bubble in a decompressed sample did not reveal significant concentration gradients. This could be a result of melt transport around the bubbles during shrinkage. The comparison between CD and SD has shown that the simulation of continuous magma ascent requires CD experiments with reasonable decompression rates. SD resulted in extensive bubble nucleation, because H2O diffusion was limited in time during the decompression steps. CD with rates ≤0.024 MPa·s-1 facilitated continuous reduction of the supersaturation by diffusive bubble growth after bubble nucleation was triggered. The bubble number densities (NV) in glass cylinder samples from CD experiments with low decompression rates are several orders of magnitude lower and the bubble diameters are bigger than in samples of SD experiments.
Projektbezogene Publikationen (Auswahl)
- (2013) Micro-FTIR imaging: An advanced method for the determination of CO2 and H2O concentration gradients in silicate glasses. Eur. J. Min. 25: 307-316
Marxer, H. and Nowak, M.
(Siehe online unter https://doi.org/10.1127/0935-1221/2013/0025-2299) - (2015) Degassing of H2O in a phonolitic melt: A closer look at decompression experiments. J. Volcanol. Geotherm. Res. 297: 109-124
Marxer, H., Bellucci, P. and Nowak, M.
(Siehe online unter https://doi.org/10.1016/j.jvolgeores.2014.11.017)
