Geoscience Reference
In-Depth Information
a
core-mantle
boundary
pressure
of
136 GPa,
which would be larger in more massive plan-
ets, the mantle can reach temperatures above the
melting point of iron forming droplets that may
founder towards the center either continuously or
episodically. Thus, it is likely that super-Earths
are differentiated. However, Elkins-Tanton and
Seager (2008) have proposed a mechanism to avoid
differentiation by reacting iron to form FeO in the
presence of enough water. The difference in radius
between differentiated and undifferentiated rocky
planets is very small. For a 10 M
E
planet the differ-
ence is 3% (Valencia
et al
., 2010), too small to be
identified given all other uncertainties, especially
composition.
rocky
super-Earths'
mantles
with
pressures
that
roughly
scale
linearly
with
mass
(e.g.
P
P
E
(
M/M
E
)) would be composed mainly of
post-perovskite, unless there is another higher-
pressure phase transition (see Figure 9.4). In
fact, currently there is a debate on whether or
not MgSiO
3
dissociates into MgO and SiO
2
at
pressures relevant to rocky super-Earths. By using
the analog NaMgF
3
suitable for compression
experiments, Umemoto
et al
. (2006) found this
transition to have a very negative Clapeyron slope
(
∼
1 TPa, relevant
to the lowermost mantles of the most massive
super-Earths. However, more recently Grocholski
et al
. (2010), have repeated the experiment with
the same material and found that instead of the
dissociation, the material transitions into yet
another phase stable to much higher pressures
and with a much lower Clapeyron slope. The
implications for the interior of super-Earths are
twofold: the pressure and hence radius at which it
happens plus the density increase associated with
it have an effect on the total planetary radius; and
the steepness of the Clayperon slope determines
whether or not mantle convection can be whole,
or layered. However, Tsuchiya and Tsuchiya
(2011) showed that when a new high-pressure
phase of SiO
2
is considered, the dissociation of
post-perovskite to oxides at
−
30 MPa
/
K) and to happen at
∼
9.3.2 Detected super-Earths
The composition of the known transiting
super-Earths is quite diverse. Here we describe
their properties. The results discussed are mostly
based on the model by Valencia
et al
. (2006, 2007,
2010), although they largely agree with those of
the other groups mentioned above. To investigate
whether or not planets are rocky, Valencia
et al
.
chose four representative compositions: pure
iron, Mercury-like (63% iron core by mass, 27%
silicate oxide mantle with no iron), Earth-like
(33% iron core, 67% silicate oxide mantle with
10% iron by mol), and a composition with no iron
(100% silicate oxides). While a pure-iron planet
(the densest possible) and a planet entirely devoid
of iron (the lightest possible) are unlikely to exist,
they demark the range of rocky compositions.
Both components are present on a planet because
the condensation temperature for iron is similar
to that of silicate and magnesium (e.g.
1 TPa is likely.
In addition, Mashimo
et al
. (2006) showed that
the transition to a virtually incompressible oxide
Gd
3
Ga
5
O
12
happened after 1.20 Mbars and sug-
gested that ''similar quasi incompressible oxide
phases composed of Si, Fe, Mg, an other elements
with relatively high natural abundances, rather
than Gd and Ga, might exist in the deep mantles
of large extrasolar rocky planets.'' It is important
to know which phases are present in planets
and their EOS for internal structure models to
include them.
∼
∼
1700
K). Figure 9.7 shows the range in M-R for rocky
planets (shaded region). Any data that falls above
the no-iron curve corresponds to a planet with
considerable amount of volatiles, with H-He,
and/or H
2
O as likely suspects.
For planets possessing an envelope of volatiles,
Valencia
et al
. (2010) added different amounts
of volatiles of H-He, H
2
O or a combination
of both, at the relevant temperature of the
planet above a nucleus of Earth-like composition
as a representative (albeit in no way certain)
composition. Of course, if this nucleus is
considered to be more iron-rich, the data for
(b) Differentiation
Another point about the in-
ternal structure is the degree of differentiation.
Because of the large gravity in super-Earths, it is
very likely that differentiation into layered struc-
ture such as the core-mantle structure on Earth
took place in super-Earths. From the gravitational
energy and energy delivered by giant impacts,
