Geoscience Reference
In-Depth Information
of light elements needs to be considered. We need
more precise experimental data on the melting
temperature of pure iron and iron-light-element
systems.
Figure 8.10 shows the temperature profile of
the core estimated by Terasaki et al . (2011) based
on the Fe-Si-O system. The upper and lower
bounds of the ICB temperature correspond to
the liquidus and solidus temperatures at ICB
pressure, respectively. The liquidus temperature
of Fe 75 O 5 S 20 at 136 GPa was determined to be
3600
Fe 75 O 5 S 20 alloy
4000
Fe
Liq
+
Liquidus
Solidus
Melting of pure Fe
3000
Liq
Fe
+
FeO
+
Liq
2000
Fe
+
FeO
+
Fe 3 S
200 K, which gives a lower bound of the
temperature at CMB (Figure 8.10). The solidus
and liquidus temperatures at 330 GPa, obtained
using extrapolation by the Simon equation, are
4380
±
1000
0
40
80
120
160
P, GPa
±
±
350 K, respectively, pro-
viding the lower and upper bounds of temperature
at the ICB for this composition. The ICB tem-
perature was estimated previously by using the
melting temperature of pure iron determined by
various procedures, and the present upper bound
of the ICB temperature of 5630
200 and 5630
Fig. 8.9 The phase and melting relations of Fe 75 O 5 S 20
alloy in the Fe-O-S system determined by Terasaki
et al . (2011). The liquidus phase is hcp-Fe, and the
second liqudus phase is FeO. Fe 3 Sisconsumedfirstat
the solidus temperature. The solidus temperature of
this system is very close to the eutectic temperature of
the Fe-Fe 3 S system, indicating that the effect of FeO in
the melting relation of the Fe-O-S system is very
small. Reproduced with permission of Elsevier.
±
350 K is lower
than the previous estimate; i.e.
170 K lower
than that (T mFe =
5800 K) estimated by the di-
amond anvil cell experiment (Ma et al ., 2004),
2007; Kamada, 2011), and the effect of FeO pres-
ence on the solidus temperature seems to be
negligible under megabar conditions.
6000
hcp-Fe (Ma et al ., 2004)
T L ad
(c) Constraints on temperature of the core based
on melting experiments The melting relations
can be used to estimate the thermal state of
the core. The temperatures at ICB and CMB
can be estimated by using the melting relations
of the iron-light-element systems. The melting
temperature of iron has been calculated to be
6700
5000
Liquidus
4000
T s ad
3000
Solidus
600 K at the ICB (e.g., Alfe et al ., 1999),
which is close to the melting temperature de-
termined by the shock experiments (Brown &
McQueen, 1980, 1986; Yoo et al ., 1993; Nguyen
& Holmes, 2004). The ICB has been assumed to be
the solid-liquid phase boundary and the melting
temperature of pure iron has been used to con-
strain the ICB temperature. However, the temper-
ature at the ICB, as estimated from the melting
temperature of pure iron, remains uncertain and
varies from 4000 to 7000K. In addition, the effect
±
CMB
ICB
350
100
150
200
250
300
Pressure, GPa
Fig. 8.10 The temperatures at the ICB and CMB
estimated based on the liquidus and solidus
temperatures of the Fe-Si-O system by Terasaki et al .
(2011). See the text in detail. T L ad and T S ad are the
adiabatic temperature profiles decompressed from the
liquidus and solidus temperatures at ICB, respectively.
Reproduced with permission of Elsevier.
Search WWH ::




Custom Search