Geoscience Reference
In-Depth Information
provide sufficiently fast rates (i.e.
500 m/year)
for the efficient partial melt extraction from the
uppermost 500-m layer of the subducting slab
(Figure 2.14a, 2.15a). The corresponding CO
2
flux
would be:
>
Infiltraction driven by
surface tension
Flux of CO
2
< 10
11
g/year.
penetraction depth ~1m
over first year
Buoyantcy-diven
porous flow
> 500 m/year
f
Porous
=
L
·
W
ϕ
·
v
Porous
ρ
Melt
C
wt
%
CO
2
/
100%,
(2.3)
where
L
≈
5 km is the width of partially molten
area along the solidus temperature (Currie
et al
.,
2004) (Figure 2.14a),
W
=
38 485 km is the ap-
proximate total length of modern subduction
zone segments (van Keken
et al
., 2011),
C
wt
%
L
~ 5 km
CO
2
=
33wt% is the CO
2
content in the melt (Shatskiy
et al
., 2013). These values yield
f
Porous
>
(a)
Buoyant melt diapir,
Rate of 2-km body
~0.5 m/year
Pressure-solution creep
of infiltrated layer
(thickness
10
16
gof
CO
2
/year (Figure 2.15b).
Surface-energy considerations (Watson, 1982;
Stevenson, 1986; Riley
et al
., 1990) suggest that
a carbonatite magma body in chemical equilib-
rium with its surroundings (overlying volatile-
poor ''dry'' mantle) will tend to dissipate by
wetting the dry grain edges of the host mate-
rial, because the dihedral angle measured at the
contact of silicate minerals and carbonatite melt
is much lower than 60 (Hunter and McKenzie,
1989; Minarik and Watson, 1995; Yoshino
et al
.,
2007). Since the capillary force is nondirectional,
this process would counteract melt segregation,
which is supported by the forces driving direc-
tional fluidmigration. High-pressure experiments
support a fast infiltration rate of carbonatite melt
on a millimetre scale (Hammouda & Laporte,
2000). However, this rate would diminish rapidly
if infiltration distance extends to geologically
relevant kilometre scale due to the increase of
diffusion distance and to the blurring of interfa-
cial energy difference at the interface with the dry
rock (Figure 2.15a). This explains why full dissi-
pation of carbonatite magma chambers into the
''dry'' mantle does not occur in reality and why
melt segregation, for example of carbonatite and
kimberlite magmas happens.
Infiltration involves chemical solution of sili-
cate at grain edges and simultaneous precipitation
of silicate crystals within the melt reservoir. To
balance the flux of the equilibrium melt into
the nonporous solid aggregate, an equal diffusive
∼
20 m)
(b)
(c)
Fig. 2.14
Schematic illustration of the formation of a
carbonatite diapir as a result of partial melting in the
CO
2
-rich uppermost layer of the subducting slab in the
transition zone. Three mechanisms of the melt
transport are involved: (1) buoyancy-driven porous flow
within partially molten slab; (2) surface tension-driven
infiltration of ''dry'' overlaying mantle;
(3) buoyancy-driven diapir ascent accompanied by
pressure-solution creep of the infiltrated layer. See text
for details.
Empirical relationships with values of
n
≈
3and
C
10 to 270 were inferred from experimental
data (Wark &Watson, 1998; Connolly
et al
., 2009;
Zhu
et al
., 2011). Considering recent experimen-
tal results on the porous flow of basaltic melt
(Connolly
et al
., 2009; Zhu
et al
., 2011) and the
lower viscosity of carbonatite melt (by 3-4 orders
of magnitude), both porous flow regimes should
≈
