 |
| Titel |
Chemical diffusion during isobaric degassing of magma |
| VerfasserIn |
Felix W. von Aulock, Ben M. Kennedy, Yan Lavallée, Sarah Henton-de Angelis, Christopher Oze, Daniel J. Morgan, Steve Clesham |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250097864
|
| Publikation (Nr.) |
 EGU/EGU2014-13484.pdf |
|
|
|
|
|
| Zusammenfassung |
| During ascent of magma, volatiles exsolve and bubbles form. Volatiles can either escape
through a permeable network of bubbles in an open system or be trapped in non-connected
pores during closed system degassing. Geochemical studies have shown that in most cases
both- open system and closed system degassing take place at the same time. During cooling
of the melt, diffusion slows down and eventually diffusional gradients get frozen in,
preserving a history of degassing and rehydration during bubble growth, bubble collapse and
crystal growth.
We present data from experiments in which natural obsidian was degassed at atmospheric
pressures at 950ºC over timescales of 3-24h. During bubble growth, a skin formed, at the
outer edge of the sample, effectively prohibiting any degassing of its interior. Diffusion
gradients were measured across the glass surrounding vesicles, and across this impermeable
skin. Water contents were analyzed with synchrotron sourced Fourier transform infrared
spectroscopy and several major, minor and trace elements were mapped using synchrotron
sourced X-ray fluorescence spectroscopy.
The samples show a dimpled surface, as well as signs of oxidation and growth of
submicroscopic crystals. Water contents around bubbles decrease in simple heating
experiments (from ~0.13 wt. % down to ~0.1 wt. %), whereas slight rehydration of the
vesicle wall can be observed when a second, cooler step at 850ºC follows the initial 950ºC.
Water gradients towards the outside of the sample decrease linearly to a minimum of ~0.045
wt. %, far below the solubility of water in melts at these temperatures. We mapped the
distribution of K, Ca, Fe, Ti, Mn, Rb, Sr, Y and Zr. Especially the trace elements show a
decrease towards the outside of the sample, whereas K, Fe, Ca and Ti generally
do not show significant partitioning between melt and gas/crystal phase. Several
effects could attribute to the distribution of these elements, such as the crystal growth
and exchange with atmospheric oxygen, and detailed models of the diffusion of
these elements will have to verify the mechanisms of elemental partitioning during
degassing
Our experiments show that even on a small scale, open system and closed system
degassing inherently coexist. This manifests itself in different elemental distribution in the
quenched glass. Water distribution gradients can be explained with diffusion during
exsolution and rehydration during cooling, however, the surface of the sample is
undersaturated in water. Some trace elements follow the same pattern, even though they
might not be considered as volatile. Therefore we suggest that chemical gradients may be
partially induced by the growth of sub-microscopic crystals and by exchange with the
atmosphere. Crystal rich, volatile poor outer skins, as produced in the experiments of this
study, have locally drastically increased viscosities and can therefore withstand higher
pressures during foaming of the interior of the sample. This self sealing of magma could be
an important process on different scales of magma degassing, from bread crust bombs to
rising magma in conduits. |
|
|
|
|
|
|