LuFe
2
O
4+x
:
from multi-ferroicity
to oxygen storage
A detailed investigation of multiferroic
LuFe
2
O
4
has shown that the layered
structure of this compound is remarkably
efficient in accommodating oxygen insertion
or desorption, thus making rare-earth ferrites
interesting candidates for applications
as oxygen sensors or solid oxygen fuel
cell oxides.
Transition metals oxides have been
generating numerous studies in materials
science for many years, owing to the
variety of their structures and properties,
some of the best-known examples being
high critical temperature superconducting
cuprates, or magnetoresistive manganites.
Multiferroic compounds,
i.e.
, materials
which exhibit two coupled ferroic
properties, also occupy a privileged place,
as the possibility to modify a property
while playing on the other opens a large
array of potential applications. The
material under study, LuFe
2
O
4
, caught
our attention because it was reported
to be
ferroelectric
at room temperature -
in relation with a charge ordering of iron
+2 and +3 species [1], and
anti-
ferromagnetic
, below 250 K [2].
In addition, the structure of this oxide
is rather unique because it combines bi-
dimensional and frustration characteristics,
originating from the stacking of triangular
[Fe
2
O
4
] and [LuO
2
] slices (Figure
➊
).
Combining different complementary
techniques, X-ray, neutron and electron
powder diffraction (XRPD, NPD and ED,
respectively), transmission electron
microscopy including high resolution
(TEM and HRTEM), thermal gravimetric
analysis (TGA), and magnetic and
electrical characterizations, we started
to investigate the impact of small changes
in the synthesis conditions upon
the structural and magnetic behaviors
of LuFe
2
O
4
samples. The observation
of small additional Bragg peaks, with
respect to the expected structure,
on the XRPD pattern, recorded at room
temperature on the CRISTAL beamline,
was hitherto confirmed by the existence
of weak satellites in the ED patterns
in some areas of the crystals [3]. The
coherence between both types of results
led us to attribute these satellites to an
incommensurate modulation, originating
from a very low oxygen excess with
respect to stoichiometric LuFe
2
O
4
, and
involving a nanoscale segregation between
more and less oxygenated areas. This first
indication of non-stoichiometry was at the
origin of a comprehensive study of oxygen
insertion - desorption in this ferrite.
TGA in controlled atmospheres, coupled
with TEM and XRPD studies, have shown
in the course of this subsequent study
that it is possible to vary the oxygen
content,
significantly
, and in a
controlled
manner, from LuFe
2
O
4
to LuFe
2
O
4.5
. The
characteristics of the successive steps are
schematized in Figure
➋
. The introduction
of oxygen in the matrix occurs through
the progressive extension of modulated
regions, followed by sliding, layer by layer,
of the [Lu] and [Fe] planes against each
other, until a new monoclinic structure
is formed. This structure (O
4.5
) is very
stable up to 700°C and, moreover,
reducing annealing (under Ar/H
2
) restores
the initial structure (O
4
). This intercalation/
de-intercalation mechanism is topotactic,
as in the battery materials, and the
compound quality is maintained despite
the relative sliding of the layers.
The cycling ability was tested with five
consecutive cycles of oxidation/ reduction,
and the storage capacity was determined
to be 1642 μmolOg
-1
, in agreement
with the O
4
to O
4.5
evolution, which
is accompanied by a rather uncommon
increase of the cell volume. The reaction
conditions at low temperatures,
the high sensitivity to oxygen, the stability
of the different phases (vs. the oxygen
content) and the structural phenomena
involved in the transitions allow one
to consider the use of this material
in systems such as sensors, solid oxygen
fuel cell oxide (SOFC) or catalysts.
Indeed, the layer sliding mechanism
is known to accommodate efficiently non-
stoichiometry and to prevent structural
collapses. In addition, the wide range
of temperature, of time and of partial
oxygen pressure, in which both valences
of iron (+2 and +3) can coexist, also
suggests that this compound could be
used as a catalyst for the degradation
of pollutants and alkanes.
The interest for the LnFe
2
O
4
type
of compounds (Ln = Y and Ho Lu)
is reinforced by this discovery,
and add another degree of freedom,
oxygen content
, as a new path
to multifunctionality.
PHYSICS AND CHEMISTRY OF CONDENSED MATTER, EARTH SCIENCES
108
SOLEIL
HIGHLIGHTS
2013
