Environmental Engineering Reference
In-Depth Information
was present, toxic reduced gases such as CO and CH
4
were also released (Ganino
and Arndt,
2009
;Svensen
et al
.,
2009
). In addition, contamination of magmas by
sediments could generate signi
fluxes of Cl, S and F when the lavas degassed
upon eruption (Sobolev
et al
.,
2009a
; Black
et al
.,
2012
).
In this paper we reappraise the quantities and types of gas released from the
two sources during the emplacement of the Siberian LIP and re-evaluate their
potential contribution to the end-Permian mass extinction.
cant
10.2 Magmatic volatiles
-
data from melt inclusions in phenocrysts
The only reliable sources of information on the magmatic volatiles in ancient volcanic
or intrusive rocks are volcanic glasses and melt or
fluid inclusions in phenocrysts.
Fresh glasses of basaltic composition are very rare, however, in continental
ood
volcanic rocks. The compositions of glasses and inclusions, including volatiles, can be
measured using modern
in situ
techniques such as electron microprobe, ion microp-
robe and laser-source inductively coupled plasma mass spectrometry (LA-ICP-MS).
Various processes can change the concentrations of volatiles in both glasses and
inclusions. The most important is degassing during magma ascent. This process
mostly affects CO
2
, which is much less soluble in the melt at crustal pressures, but it
also may affect H
2
O if degassing takes place at shallow depths. The most resistant
to degassing are Cl, F and S, which remain relatively soluble in basaltic melts at low
pressures close to the surface. Because inclusions in phenocrysts normally are
trapped and partially isolated in relatively deep magma chambers well before final
eruption, melt inclusions are usually less degassed than glasses and better retain the
original volatile contents (Sobolev,
1996
). However, certain elements may
exchange by diffusion between the included melt and enclosing crystal and magma.
Portnyagin
et al
.(
2008
) showed that the H
2
O contents of melt inclusions could be
signi
ed by diffusion of hydrogen either into or out of the inclusion.
The timescale of this process varies from hours to minutes. As temperature
decreases in slowly cooled olivine cumulates, Fe
cantly modi
Mg exchange between included
melt and host olivine results in Fe loss from the melt (Sobolev and Danyushevsky,
1994
). This, in turn, decreases the sulfur solubility and leads to precipitation of a
sul
-
de melt, thus decreasing the sulfur content of the included melt (Danyushevsky
et al
.,
2002
). In addition, CO
2
can be stored in the gas bubbles in inclusions, thus
compromising data on the C concentration in the glass (e.g. Bucholz
et al
.,
2013
).
10.2.1 Samples and methods
As shown in
Figure 10.1
, our samples cover most of the stratigraphic sequence and
almost the entire period of eruption of lavas of the Siberian LIP. Here we report
