When Boehler [1986] presented his first data on 
of iron,
he predicted a new phase above 200 GPa by extrapolation, as mentioned
above. When extrapolated, his data on 
, then measured to 42 GPa,
appeared to intersect the preliminary shock wave data of
Brown and McQueen [1982, 1986], who had reported a
solid--solid (s-
) transition at 200 GPa and 
. It was clear that the melting curve given by
Boehler's solid (hcp)-liquid boundary could not be below
the B&M s-
boundary. Not only that, but which two solid phases
could have a boundary at 200 GPa? The fcc 
phase is locked
off at low pressures by the join between the 
-
-
triple point and the 
-
-liquid
triple point at 100 GPa found by Boehler. Therefore,
Boehler reasoned there is a new phase in the phase
diagram above 200 GPa, which he named 
.
At the 1989 Los Alamos ironworkers convention,
there was much discussion of what the structure of this 
phase might be, and three Livermore theorists [ Ross et al., 1990]
suggested that the 
phase could have a bcc structure.
This suggestion was based on analogies to other transition
elements in which the bcc structure is in stable equilibrium
with the liquid. By the time of the 1993 Ironworkers convention,
many theorists had tested the possibility of a bcc phase with
their first principles methods. Of the five theorists who
participated in that meeting, four found bcc to be unstable
at high pressure. One [ Matsui, 1994] found from his
molecular dynamics calculations that there is an hcp-bcc boundary
at 
where 
.
But Sherman [1994], using a variation of the LAPW method,
found bcc to be unstable under compression. Similarly, Stixrude
and Cohen [1994, 1995] and Stixrude et al. [1994] found
bcc to be mechanically unstable at all pressures. Moriarty [1994]
also found the bcc structure to be mechanically unstable
at all high pressures, because it is associated with imaginary
phonons and an imaginary 
elastic constant.
Thus the high pressure phase
should not be bcc.
But neither can it be hcp,
if the t.p. at about
190 GPa is secure (Figure 1),
and the new 
data
of Gallagher and Ahrens at
200 GPa make this t.p. robust.
Moriarty [1994] suggested that
might be fcc.
Stixrude and
Cohen [1994] found that
the Helmholtz energy curves for the
hcp and fcc phases are nearly
coincident at all pressures,
so there is a possibility
that fcc could have a
smaller Gibbs energy than hcp
at inner core pressures.
Following the suggestion of Moriarty [1994]
that the inner core region could be dominated by fcc,
Isaak et al. [1994] found that the density of fcc iron
at conditions of the inner core region probably has a value between
that of hcp and that of the PREM inner core. They used a Birch-Murnaghan
3rd order EoS and the Morse potential EoS for the calculation, where
is the isothermal bulk
modulus of fcc at ambient conditions; they found
.
This value of 
was estimated from the value of
at 1430 K measured by Stassis [1994] by
neutron diffraction data and a subsequent lattice dynamic analysis.
The value of 
for fcc iron appears to be close to that of
for hcp iron (193 GPa with 
)
[ Jephcoat et al., 1986], but the value of 
is
smaller for fcc than for hcp (where 
).
The problem of accounting for a fourth phase, 
,
in first principles calculations of iron will be formidable.
And there may be yet another phase, the fifth, occurring
at lower pressures. Saxena et al. [1994] announced that
they had uncovered a phase boundary approximately
parallel to (but at temperatures less than) the join between
the two triple points: 
-
-
and 
-
-
(in the region of 2000 K and
50--80 GPa). They named this phase (of unknown crystallographic
structure) 
.
A similar new solid-solid boundary,
slightly displaced from
the s-
boundary
of Saxena et al. [1994],
was reported by Boehler [1994].
The existence of a 
phase
would require that
the 
curve
above the triple point (which is
at 100 GPa, according to Boehler)
be the boundary between the liquid
and the 
phases.
There is currently a substantial
effort at the Brookhaven
Synchrotron Radiation Facility
to measure the crystallographic structure
of iron in the neighborhood
of the claimed 
phase
[ Yoo et al., 1994b]. As of this
writing, the results of Yoo et al. [1994b]
indicate that extra phases
beyond the standard 
-
-
phases do
exist, but have not yet been identified.