Studies on structural, magnetic and thermal properties of xFe 2TiO4-(1-x)Fe3O4 (0≤x≤1) pseudo-binary system

Article abstract of DOI:10.1016/j.jmmm.2011.12.012

The xFe₂TiO₄-(1-x)Fe₃O₄ pseudo-binary systems (0≤x≤1) of ulv?spinel component were synthesized by solid-state reaction between ulv?spinel Fe ₂TiO₄ precursors and commercial Fe₃O

Full text of DOI:10.1016/j.jmmm.2011.12.012

Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
Contents lists available at SciVerse ScienceDirect
Journal of Magnetism and Magnetic Materials
journal homepage: www.elsevier.com/locate/jmmm
Studies on Structural, Magnetic and Thermal Properties of
xFe2TiO4-(1ꢀx)Fe3O4 (0rxr1) Pseudo-binary System
n
a
b
c
b
Monica Sorescu a, , Tianhong Xu , Adam Wise , Marina Dıaz-Michelena , Michael E. McHenry
´
a Department of Physics, Duquesne University, Fisher Hall, Pittsburgh, PA 15282, USA
b Department of Materials Science and Engineering, Carnegie Mellon University, Roberts Hall, Pittsburgh, PA 15213, USA
c Space Programs and Space Sciences Department, Instituto Nacional de Te´cnica Aeroespacial (INTA), Madrid 28850, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The xFe2TiO4-(1ꢀx)Fe3O4 pseudo-binary systems (0rxr1) of ulvo¨spinel component were synthesized
by solid-state reaction between ulvo¨spinel Fe2TiO4 precursors and commercial Fe3O4 powders in
stochiometric proportions. Crystalline structures were determined by X-ray powder diffraction (XRD)
and it was found that the as-obtained titanomagnetites maintain an inverse spinel structure. The lattice
parameter a of synthesized titanomagnetite increases linearly with the increase in the ulvo¨spinel
component. 57Fe room temperature Mo¨ssbauer spectra were employed to evaluate the magnetic
properties and cation distribution. The hyperfine magnetic field is observed to decrease with increasing
Fe2TiO4 component. The fraction of Fe2þ in both tetrahedral and octahedral sites increases with the
increase in Ti4þ content, due to the substitution and reduction of Fe3þ by Ti4þ that maintains the
charge balance in the spinel structure. For x in the range of 0 rxr0.4, the solid solution is
ferrimagnetic at room temperature. However, it shows weak ferrimagnetic and paramagnetic behavior
for x in the range of 0.4oxr0.7. When x40.70, it only shows paramagnetic behavior, with the
appearance of quadrupole doublets in the Mo¨ssbauer spectra. Simultaneous differential scanning
calorimetry and thermogravimetric analysis (DSC–TGA) studies showed that magnetite is not stable,
and thermal decomposition of magnetite occurs with weight losses accompanying with exothermic
processes under heat treatment in inert atmosphere.
Received 18 August 2011
Received in revised form
15 November 2011
Available online 9 December 2011
Keywords:
Titanomagnetite
Mars
X-ray powder diffraction
Mo¨ssbauer spectroscopy
Simultaneous DSC–TGA
& 2011 Elsevier B.V. All rights reserved.
1. Introduction
Important components among the abundant mineral types on
the Mars are ferrites and titanomagnetites, and those materials
In basalts, Fe–Ti spinels (titanomagnetite) are the carriers of
the natural remanent magnetization of Mars. Studies of the
magnetic properties of such rocks have provided a wealth of
information on the magnetic history of Mars. The further studies
of planetary magnetism provide insight into the validity of
various models of geological evolution of planets and satellites
and thus of the Solar System. The Mars Global Surveyor mission’s
results showed the first unambiguous measurements of the
magnetic field on Mars [1]. These results showed that Mars
presently has no active geological dynamo and its global magnetic
field is of crustal origin even though studies of Martian meteorites
indicate that fields as large as 3000 nT may have once existed on
the surface of Mars. Mappings of the Martian magnetic field
reveal anomalous regions of intense magnetization [2] distributed
in ‘‘stripes’’ of alternating polarity.
usually contain medium Ti content. Due to their abundance on
the Martian surface, studying the formation of xFe2TiO4-
(1ꢀx)Fe3O4 solid solutions with different amounts of Ti content
and their magnetic properties may be important in understanding
the geomagnetism of Mars. In order to explain the remanent
magnetic state on Mars it is necessary to postulate abundant
materials with large remanent magnetization. This in turn
requires monodomain magnets, with sufficiently large Neel tem-
peratures with sufficient pinning of the remanent state to avoid
thermally activated switching. Spinodal decomposition in an
asymmetric miscibility gap in the titanomagnetite system can
explain these features [3,4].
Titanomagnetites are solid solutions of magnetite and ulvo¨-
spinel (Fe2TiO4), in which a fraction of Fe3þ ions on the octahe-
dral sites of magnetite are substituted by Ti4þ ions. When a Fe3þ
is substituted by a Ti4þ ion, a Fe3þ ion on either tetrahedral or
octahedral site must convert to a Fe2þ in order to maintain charge
balance. It has been reported that Ti4þ ions have strong pre-
ference for octahedral sites in the titanomagnetites [5]. However,
occupation of Ti4þ ions on tetrahedral sites was also found to be a
n Corresponding author. Tel.: þ1 412 396 4166; fax: þ1 412 396 4829.
E-mail address: sorescu@duq.edu (M. Sorescu).
0304-8853/$ - see front matter & 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmmm.2011.12.012

1454
M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
possible choice in Fe–Ni–Ti, Li–Mn–Ti and Li–Cr–Ti spinels [6–8].
The ordering of these cations is still uncertain; depending on the
molar fraction x and cation distribution on the tetrahedral and
octahedral sites, the solid solutions xFe2TiO4-(1ꢀx)Fe3O4 can
exhibit ferrimagnetic, antiferromagnetic, spin glass and paramag-
netic behaviors with different Ti4þ contents.
commercial Fe3O4 powder in different proportions to produce the
series studied here. Each composition was ball-milled for another
10 min to reduce the particle size and homogenize the powder
mixture, and then pressed into pellets of 0.5 in in diameter and
ꢁ0.15 in thick. The pellets were placed in 1010 steel crucibles
into an Ar furnace at 950 1C for a bake of 60 h. Pellets were cooled
in furnace under the Ar atmosphere to 400 1C, then removed to
air-cool to room temperature.
Mo¨ssbauer spectroscopy has proved to be a valuable tool in
geology giving information about various aspects of iron miner-
alogy that in turn can provide insight into the magnetic history of
minerals on the Mars. Recently, the cation distributions in
titanomagnetites have been widely investigated, and several
models have been proposed to describe the Fe3þ ions on tetra-
hedral and octahedral sites as a function of Ti content [9–13]. So
far, eight models were used to describe the relationship between
cation distribution and molar fraction of ulvo¨spinel component
[14]. However, these models are based on the saturation magne-
tization measurements performed at low temperature such as
77 K or Mo¨ssbauer spectroscopy measurements at 4.2 K under
strong external magnetic field, and the de-convolution of the
Fe3þ sextet on the tetrahedral and octahedral sites is not
completely effective and conclusive. Recently, X-ray magnetic
circular dichroism (XMCD) was also employed to study the Fe site
occupancy in magnetite-ulvo¨spinel solid solutions, while this
technique is only suited to probe the surface with a depth of
ꢁ4.5 nm [14]. To the best of our knowledge, no cation distribu-
tion in titanomagnetite has been successfully investigated by
room temperature Mo¨ssbauer spectroscopy without strong exter-
nal magnetic field.
Recently, thermal analysis was employed to study the oxida-
tion behavior of titanomagnetite powders [15–17]. However,
those experiments were only conducted under oxidation atmo-
sphere to study the oxidation behavior of Fe2þ either located in
octahedral B or tetrahedral A sites. No thermal behavior of the
xFe2TiO4-(1ꢀx)Fe3O4 pseudo-binary systems under inert atmo-
sphere has been studied. The simultaneous DSC–TGA investiga-
tion under inert atmosphere on the solid solution systems may
provide some useful information about the formation process of
titanomagnetite.
The as-synthesized Fe2TiO4–Fe3O4 solid solutions were pow-
dered for phase, magnetic and thermal examinations. The X-ray
powder diffraction (XRD) patterns of the solid solutions were
obtained using a PANalytic X’Pert Pro MPO diffractometer with
˚
CuK radiation (45 kV/40 mA,
on the diffracted side. A silicon-strip detector called X’cellerator
l¼1.54187 A) with a nickel filter
a
was used. The scanning range was 10–801 (2y) with a step size of
0.021.
Room temperature transmission Mo¨ssbauer spectra were
recorded using a MS-1200 constant acceleration spectrometer
with a 10 mCi 57Co source diffused in Rh matrix. Least-squares
fittings of the Mo¨ssbauer spectra were performed with the
NORMOS program.
Simultaneous DSC–TGA experiments were performed using a
Netzsch Model STA 449 F3 Jupiter instrument with a Silicon
Carbide (SiC) furnace. Samples were contained in an alumina
crucible with an alumina lid. Series of experiments were per-
formed using 2072 mg sample size. The atmosphere consisted of
flowing protective argon gas at a rate of 50 ml/min. DSC and TGA
curves were obtained by heating samples from room temperature
to 800 1C with a ramp rate of 10 1C/min. Both DSC and TGA curves
were corrected by subtraction of a baseline which was run under
identical conditions as the DSC–TGA measurement with residue
of samples in the crucible. The Netzsch Proteus Thermal Analysis
software was used for DSC and TGA data analysis.
3. Results and Discussion
3.1. X-ray Powder Diffraction
This work reports the successful synthesis of solid solutions
xFe2TiO4-(1ꢀx)Fe3O4 in whole ranges with 0rxr1 by use of the
solid state reaction. The solid state reactions between synthesized
Fe2TiO4 and commercial Fe3O4 powders were performed in
similar conditions as expected to occur in Martian crust. Magnetic
properties of these as-obtained materials were investigated
through room temperature Mo¨ssbauer spectroscopy. The varia-
tions in magnetic properties of the solid solution as a function of
Fe2TiO4 molar fraction x were investigated, and the Fe cation
distribution in tetrahedral and octahedral sites was derived from
the overall oxidation state of Fe ions on the tetrahedral site. The
thermal behavior of as-obtained solid solutions under inert atmo-
sphere, which depends on the Fe2TiO4 component, was also
reported. These are important for future interpretation of mag-
netic and structural data collected on Mars.
Fig. 1 shows the X-ray powder diffraction of the xFe2TiO4-
(1ꢀx)Fe3O4 with different Fe2TiO4 molar fractions x in the range
of 0rxr1. As expected, Fig. 1a is a typical powder diffraction
pattern of single-phase cubic spinel magnetite. The XRD of the
synthesized Fe2TiO4 materials was shown in Fig. 1g, and verified
to be the spinel phase along with less than 3% impurity of
ilmenite (FeTiO3) phase, which is a typical phase approaching
ulvo¨spinel. The Fe2TiO4 spinel phase forms according to the
reaction
3TiO2 þ2Fe2O3 þ2Fe ¼ 3Fe2TiO4
ð1Þ
The XRD patterns show that the synthesized xFe2TiO4-
(1ꢀx)Fe3O4 solid solutions with x¼0.4, 0.6, 0.7 and 0.75 are cubic
spinels. Absence of any other phases confirms that a complete
solid solution has been formed between Fe3O4 and Fe2TiO4,
except for x¼0.6, where traces of FeO were detected. However,
for x¼0.65, a pure spinel cubic phase was not achieved (Fig. 1d).
Other phases, such as Fe2O3, Fe2TiO5 and FeTiO3, were detected
along with the spinel structure, suggesting that this molar
fraction may not be suitable to obtain the solid solution with
single-phase of cubic spinel. The detected three side phases may
also result from exsolution when the formed solid solution
becomes unstable under cooling process, and early spinodal
decomposition occurs [3]. It has been generally reported that
the miscibility gap in the Fe2TiO4–Fe3O4 solid solution series leads
to the development of exsolution intergrowths in titanomagne-
tites of intermediate composition [18,19], and the spinodal
2. Experimental
The xFe2TiO4-(1ꢀx)Fe3O4 (0rxr1) series were produced by
solid state synthesis. Fe2TiO4 was prepared from Fe2O3 and TiO2
mixed in stoichiometric proportions with 5.4% excess Fe sponge.
Components were ground with a mortar and pestle, pressed into
pellets, loaded into Fe crucibles and heated to 1050 1C for 60 h in
an Ar atmosphere to prevent oxidation.
The Fe2TiO4 precursor pellets were SPEX ball-milled to pro-
duce powders of an appropriate size for mixing with commercial
Fe3O4 powder. This ball-milled Fe2TiO4 powder was mixed with

M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
1455
Fig. 2. Lattice parameter a of the xFe2TiO4-(1ꢀx)Fe3O4 (0rxr1) pseudo-binary
system as a function of the Fe2TiO4 molar fraction of x.
x¼0.65 is off the curve to some extent, consistent with the
measurable amounts of hematite, Fe2TiO5, FeTiO3 impurities and
possible magnetite phase, respectively. The relationship between
lattice parameter a and Fe2TiO4 content is similar to that reported
in the literatures with S-shape curve; the small difference may
arise from the different preparation methods and the formation of
a non-stoichiometric solid solution [24,25]. The decrease in the
relative amount of Fe3O4 and increasing Fe2TiO4 content results in
a larger lattice parameter a of the solid solution xFe2TiO4-
(1ꢀx)Fe3O4. The linear change in lattice parameter a is induced
by the Ti substitution of Fe ions and the reduction of Fe3þ to
maintain charge balance in the formed solid solutions.
The spinel structure has a space group of Fd3m. This space
group requires equivalence among tetrahedral sites and among
octahedral sites. For spinel with different types of cations occur-
ring together at the same sites, a random distribution of the
cations over equivalent sites is required, otherwise the symmetry
will be reduced if such equivalent site is violated and the cations
order in various ways [24]. All of the XRD patterns are consistent
with the Fd3m space group, no extra reflection and splitting of
spinel peak was observed, indicating the Ti4þ cations are ran-
domly distributed in the formed solid solutions with different
molar fraction x of ulvo¨spinel component.
Fig. 1. XRD patterns of the xFe2TiO4-(1ꢀx)Fe3O4 (0rxr1) pseudo-binary system
with molar fraction x (a) x¼0.0, (b) x¼0.4, (c) x¼0.6, (d) x¼0.65, (e) x¼0.70,
(f) x¼0.75 and (g) x¼1.0.
decomposition in titanomagnetite was also observed from the
electron microscopy studies [2,20]. This asymmetry in the mis-
cibility gap is not predicted from regular solution theory and must
be further understood in the future. However, the X-ray powder
diffraction peaks of Fe3O4 from spinodal decomposition were not
observed, probably due to the strong overlap in diffraction peaks
between the Fe3O4 and the as-obtained Ti-rich spinel structures.
The broadening of the XRD peaks is consistent with a fine
microstructure characteristic of spinodal decomposition.
3.2. Mo¨ssbauer Spectroscopy
Both magnetite and ulvo¨spinel have a cubic inverse spinel
structure, with the structure formula of (Fe3þ)[Fe2þFe3þ]O4 and
(Fe2þ)[Fe2þTi4þ]O4, respectively. The round bracket stands for
tetrahedral site and square bracket for octahedral site, respec-
tively. Cation distribution in solid solution titanomagnetites is
essential to understand the complicated magnetic behavior of
these minerals with remanent magnetism in rocks, which
includes the magnetocrystalline anisotropy and spin canting
[13]. The substitution of non-magnetic Ti4þ can have similar
effects to the substituion of non-magnetic Al3þ [26] and Zn2þ in
Ni–Zn ferrites [27,28]. The preferential breaking of A–B super-
exchange bonds can stabilize Yafet–Kittel triangular spin struc-
tures causing dipole moments to have to switch against stronger
exchange interactions. This manifests itself in non-saturating
magnetization curves up to large fields [4,26] causing difficulty
inferring cation occupancies from saturation magnetizations.
Magnetic behavior can further be complicated by spatially
Compared to the XRD patterns of the as-obtained spinel
materials, it was found that the diffraction peaks shift to lower
2y angles systematically with the increase in the molar fraction x
of Fe2TiO4, owing to the difference in radius between Fe3þ/Fe2þ
and Ti4þ. The increase in the lattice parameter a of the as-
obtained spinel materials with the increase in the Ti4þ content
is also suggested. Fig. 2 shows variation of the lattice parameter a,
which was extracted from the Rietveld structural refinement of
the XRD patterns, as a function of the Fe2TiO4 molar fraction x. It
was found that there is a linear relationship between a and x, a
˚
˚
increases from 8.3894 A to 8.5410 A for x¼0 and x¼1, respec-
tively. The linear relationship follows the Ve´gard’s rule [21,22],
such that the lattice parameter a of the formed solid solution has
a weighted average of that of the two end members [23]. It has to
be noted that the lattice parameter a of the solid solution with

1456
M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
inhomogeneous cation distributions caused by spinodal decom-
position [29] or diffusion for surfaces or interfaces [28].
Table 1
Mo¨ssbauer parameters for the xFe2TiO4-(1ꢀx)Fe3O4 (0.0rxr1) pseudo-binary
system.
The three Mo¨ssbauer parameters, namely isomer shift (I.S.,
relative to a-Fe), quadrupole shift/splitting (Q.S.) and hyperfine
magnetic field (B), can provide important information about a
particular compound with an atomic resolution scale. Especially,
the isomer shift provides direct information about the electron
density at nucleus.
Magnetite (Fe3O4) is an oxide with the inverse spinel structure,
which has one Fe3þ ion on the tetrahedral site and two Fe ions,
with a total oxidation state of þ5, on the octahedral site. Fig. 3a
shows the room temperature Mo¨ssbauer spectrum of magnetite.
It could be resolved by considering two Zeeman sextets (Table 1),
corresponding to the two magnetic sublattices present in the
sample. The tetrahedral site is represented by a sextet with a
magnetic hyperfine field of 49.22 T and isomer shift of 0.059 mm/s,
while the octahedral site is described by a six-line pattern having a
hyperfine magnetic field of 46.20 T and isomer shift of 0.472 mm/s.
Lowering of the hyperfine magnetic field value of the second
sublattice is related to the presence of Fe2þ ions at the octahedral
sites. The site populations of the tetrahedral and the octahedral
X
I.S. (
d
, mm/s)
Q.S. (mm/s)
B (T)
Abundance (%)
x¼0.0
x¼0.40
0.059
0.472
0.191
0.071
0.550
0.489
0.318
0.341
0.417
0.700
0.231
0.332
0.360
0.470
0.252
0.332
0.373
0.387
0.753
0.772
0.768
0.793
0.904
0.903
0.024
49.22
46.20
46.75
44.29
41.99
36.71
38.45
33.25
27.63
/
34.6
65.4
27.7
11.1
30.6
30.6
29.0
11.8
33.2
26.0
12.8
22.8
6.7
0.047
ꢀ0.039
0.360
ꢀ0.155
ꢀ0.222
0.076
x¼0.60
x¼0.65
0.088
ꢀ0.022
0.778
ꢀ0.012
ꢀ0.082
0.044
50.71
46.80
44.19
40.72
/
0.176
42.0
15.7
33.4
9.8
0.558
x¼0.70
x¼0.75
0.033
27.83
22.22
19.48
/
0.072
ꢀ0.150
1.211
2.188
1.378
1.680
36.5
20.3
8.1
/
/
38.9
53.0
50.3
49.7
70.5
/
x¼1.0
1.223
1.976
70.02
/
/
Error
70.01
71.00
sites are 34.6% and 65.4%, respectively. The ratio of the population
of Fe ions on tetrahedral and octahedral sites is close to the
theoretical value of 1:2, indicating a high degree of stoichiometry
of the initial magnetite material [30]. The spectrum also indicates
that magnetite does not contain any Fe2þ in tetrahedral sites, while
it contains both Fe3þ and Fe2þ in octahedral sites. As an equal
number of Fe3þ and Fe2þ are present at octahedral sites, an average
oxidation state of þ2.5 can be assumed for Fe ions in octahedral
sites, due to the electron hopping between Fe3þ and Fe2þ
.
Fig. 3g shows the room temperature Mo¨ssbauer spectrum of
the as-synthesized Fe2TiO4. The spectrum of Fe2TiO4 is very
different from that of magnetite; instead of two sextets, only
one broad doublet is visible. The least-square fitting shows that
the broad doublet is composed of two quadrupole doublets. The
corresponding values of isomer shift (ꢁ0.90 mm/s) and quadru-
pole splitting (1.2rQ.S.r2.0 mm/s) for these two doublets are
much larger than those of Fe3þ on the tetrahedral site and Fe2.5þ
on the octahedral site of magnetite, indicating that Fe ions exist
only as Fe2þ ions in the spinel Fe2TiO4. This is in good agreement
with the charge balance, which implies that Ti ions exist as Ti4þ
and Fe ions exist as Fe2þ. Though a very small portion of
impurities of FeTiO3 was detected from XRD measurement, the
contribution of FeTiO3 to the overall magnetic properties is also
from Fe2þ ions. On the other hand, the percentage of FeTiO3 was
found to be less than 3%, therefore, the two doublets can be
mainly considered from the contribution of Fe2þ in Fe2TiO4. The
two doublets in the Mo¨ssbauer spectrum of Fe2TiO4 can be
explained in terms of the Fe2þ ions on both tetrahedral site and
octahedral sites [31]. The percentage of these two doublets from
least-square fitting follows the almost equal distribution of Fe2þ
ions on the tetrahedral (Q.S.¼1.223 mm/s) and octahedral
(Q.S.¼1.976 mm/s) sites, with the percentage of 50.3% and
49.7%, respectively. The existence of small different percentages
is reasonable since the field gradient values at the Fe2þ ions are
different because of the partial inversion of the synthesized
crystal as suggested from Tanaka et al. [32,33] or by other
structural reasons, such as vacancies in the neighborhood and
impurities. The almost equal amount of Fe2þ on tetrahedral and
Fig. 3. Mo¨ssbauer spectra of the xFe2TiO4-(1ꢀx)Fe3O4 (0rxr1) pseudo-binary
system with molar fraction
x
(a) x¼0.0, (b) x¼0.4, (c) x¼0.6, (d) x¼0.65,
(e) x¼0.70, (f) x¼0.75 and (g) x¼1.0.

M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
1457
octahedral site also suggests that Ti4þ ions are located on the
octahedral site of Fe2TiO4. It was also reported by Vanleerberghe
and Vandenberghe [34] that no crystallographic ordering
between Ti4þ and Fe2þ ions on octahedral site was observed,
which results in a whole range of possible distributions of the
Fe2þ surroundings and gives rise to distributions in the quadru-
pole interaction for both sites. Therefore, it is not surprising to
find only one paramagnetic broad doublet spectra as show in
Fig. 3g. Only one broad doublet is observed for the synthesized
Fe2TiO4 sample, suggesting it is paramagnetic at room tempera-
ture, which is in good agreement with the reported results that
Fe2TiO4 is paramagnetic at room temperature and antiferromag-
netic at very low temperature [35]. From the above Mo¨ssbauer
analysis of magnetite and Fe2TiO4, the cation distribution is well
established for these two end members. In order to determine the
cation distribution in solid solution xFe2TiO4-(1ꢀx)Fe3O4 with
different x values, the room temperature Mo¨ssbauer spectra were
also recorded.
with the highest hyperfine magnetic fields of 38.45 T and 33.25 T,
can be assigned to Fe ions on tetrahedral sites. The percentage of
Fe ions on the tetrahedral sites is the sum of the populations of Fe
ions with higher hyperfine magnetic fields, which is about 40.8%
and very close to the theoretical value of 41.66% as calculated
from the (Fe1)[Fe1.4Ti0.6]O4 formula. The other sextet can be
assigned to Fe ions on octahedral sites. The decrease in the
hyperfine field strength of Fe ions located on tetrahedral sites
suggests that part of Fe3þ ions were reduced due to the substitu-
tion between Ti4þ and Fe ions. The doublet has a population of
26.0%, and with isomer shift and quadrupole splitting values of
0.700 and 0.778 mm/s, which can be assigned to Fe ions on
octahedral sites. The appearance of the quadrupole splitting
doublet suggests that more Fe3þ ions are reduced by the
introduction of more Ti4þ into the solid solutions.
Fig. 3d shows the transmission Mo¨ssbauer spectrum of the
solid solution xFe2TiO4-(1ꢀx)Fe3O4 with x¼0.65, corresponding
to Fe2.35Ti0.65O4. The Mo¨ssbauer spectrum is composed of super-
imposed four magnetic hyperfine sextets and one doublet. The
corresponding hyperfine magnetic fields for the sextets are 50.71,
46.80, 44.19 and 40.72 T, respectively, with areal percentages of
12.8%, 22.8%, 6.7% and 42.0%, respectively. The presence of
measurable amounts of hematite, FeTiO3 and Fe2TiO5 also makes
it more difficult to quantify the Fe cation distribution on tetra-
hedral and octahedral sites.
Fig. 3e shows the transmission Mo¨ssbauer spectrum of the
solid solution xFe2TiO4-(1ꢀx)Fe3O4 with x¼0.7, corresponding to
Fe2.3Ti0.7O4. The Mo¨ssbauer spectrum is composed of three super-
imposed magnetic hyperfine sextets and one doublet. The corre-
sponding hyperfine magnetic fields for the three sextets are 27.83,
22.22 and 19.48 T, respectively, with areal percentages of 33.4%,
9.8%, 36.5%, respectively. Compared to the solid solution xFe2-
TiO4-(1ꢀx)Fe3O4 with x¼0.6 and 0.4, the hyperfine magnetic
field strengths were found to be dramatically smaller. The sextets
with the two highest hyperfine magnetic fields of 27.83 T and
22.22 T, and isomer shifts of 0.332 and 0.373 mm/s, respectively,
can be assigned to Fe ions on tetrahedral site. The other sextet and
the doublet can be assigned to Fe ions on B sites. The percentage
of Fe ion on the tetrahedral site is about 43.2%, which is very close
to the ideal spinel structure and theoretical value of 43.38% as
calculated from the (Fe1)[Fe1.3Ti0.7]O4 formula. Furthermore, this
also suggests that Ti4þ remains on the octahedral site. However,
some of Fe3þ ions on tetrahedral site are reduced, which results
in a much smaller hyperfine magnetic field strength. Analysis of
the values of isomer shift and quadrupole splitting of the doublet,
with the value of 0.753 and 1.211 mm/s, suggests Fe ions are
further reduced to Fe2þ on octahedral site at this Ti4þ content. In
order to keep the charge balance in the spinel structure, the
percentage of Fe ions with the reduced oxidation state increases
with the increase in the molar fraction x of Fe2TiO4, which can be
confirmed from the increase in areal percentage of the doublet.
Fig. 3f shows the room temperature Mo¨ssbauer spectrum of
the solid solution xFe2TiO4-(1ꢀx)Fe3O4 with x¼0.75, correspond-
ing to Fe2.25Ti0.75O4. Different from other samples and similar to
the Fe2TiO4, the Mo¨ssbauer spectrum of the solid solution is
composed of one broad asymmetric doublet, which can be de-
convoluted to three doublets, and with areal percentages of 8.1%,
38.9%, 53.0%, respectively; no sextet was observed. Analysis of the
values of isomer shift and quadrupole splitting of the doublets can
determine the Fe cation distribution on tetrahedral site and
octahedral site. The doublets with the isomer shift values of
0.772 mm/s and 0.768 mm/s can be assigned to Fe ions located on
tetrahedral site, with a population of 47.0%, close to the theore-
tical value (44.44%) with formula of (Fe1)[Fe1.25Ti0.75]O4. The
other doublet, with isomer shift and quadrupole splitting values
of 0.793 and 1.680 mm/s can be assigned to Fe ions on octahedral
Fig. 3b shows the transmission Mo¨ssbauer spectrum of the
solid solution xFe2TiO4-(1ꢀx)Fe3O4 with x¼0.4, corresponding to
Fe2.6Ti0.4O4. Compared to the magnetic properties of original
magnetite and Fe2TiO4 samples, the formed solid solution showed
completely different magnetic properties, though it still main-
tains the spinel structure, as identified from XRD measurement.
Local variations in hyperfine fields produce broad lines in the
Mo¨ssbauer spectrum. Least-square fitting shows that the
Mo¨ssbauer spectrum for this sample is composed of superim-
posed four magnetic hyperfine sextets, and the overall hyperfine
magnetic field decreases with the introduction of Fe2TiO4, with
the values of 46.75, 44.29, 41.99 and 36.71 T, respectively. The
two sextets, with the hyperfine magnetic fields of 46.75 and
44.29 T, respectively, can be assigned to Fe ions on tetrahedral
sites. The other two sextets with hyperfine field of 41.99 and
36.71 T can be assigned to Fe ions on octahedral sites. The isomer
shift of Fe ion on tetrahedral site is smaller than that on
octahedral site. Population of Fe ions on the tetrahedral sites, in
other words, the sum of the two sextets with higher magnetic
fields, is about 38.8%, which is very close to the theoretical value
of 38.45% as calculated from the (Fe1)[Fe1.6Ti0.4]O4 formula,
suggesting that Ti4þ remains at octahedral site. This is consistent
with Mo¨ssbauer result of Fe2TiO4 that Ti4þ occupies octahedral
sites. Similar to the Mo¨ssbauer spectrum of Fe3O4, the rapid
electron hopping between Fe2þ and Fe3þ on the octahedral sites
makes it very difficult to completely separate the Fe2þ and Fe3þ
cations on octahedral sites, although the spectrum can be fitted
with two sextets with smaller hyperfine magnetic fields. The
existence of two sextets for Fe ion on tetrahedral site suggests
that part of Fe3þ was reduced as well, with the sign of decrease in
the strength of hyperfine magnetic field and increase in isomer
shift, respectively. The Mo¨ssbauer result shows the cation dis-
tribution in this synthesized solid solution is following the model
reported by O’Reilly et al [11]. The hyperfine magnetic field
strength as a function of Ti content (0rxr0.5) was also studied
by use of room temperature Mo¨ssbauer spectra [36–38] and it
was reported that the outer sextet of the xFe2TiO4-(1ꢀx)Fe3O4
solid solution with x¼0.4 has a hyperfine magnetic field of 45.5 T;
our results are in great agreement with the reported values.
Fig. 3c shows the transmission Mo¨ssbauer spectrum of the
solid solution xFe2TiO4-(1ꢀx)Fe3O4 with x¼0.6, corresponding to
Fe2.4Ti0.6O4. Similar to that of x¼0.4, the formed solid solution
showed completely different magnetic properties from those of
original magnetite and Fe2TiO4 samples. The Mo¨ssbauer spectrum
is composed of three superimposed magnetic hyperfine sextets
and one doublet. The corresponding hyperfine magnetic fields for
the sextets are 38.45, 33.25 and 27.63 T, respectively, with areal
percentages of 29.0%, 11.8%, 33.2%, respectively. The two sextets,

1458
M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
sites. Considering this Fe2TiO4 molar fraction of x¼0.75 and the
charge balance in the spinel structures, some portion of Fe ions on
tetrahedral sites should still maintain Fe3þ character. However, no
sextet corresponding to Fe3þ on tetrahedral site was observed. Most
of the Fe ions exist as Fe2þ on tetrahedral sites and some are Fe3þ
,
however, the solid solution only shows paramagnetic behavior at
room temperature, indicating rapid electron hopping between Fe2þ
and Fe3þ ions on tetrahedral site as well. The paramagnetic proper-
ties are probably due to the dominating amount of Fe2þ ions in the
solid solution xFe2TiO4-(1ꢀx)Fe3O4 with molar fraction x¼0.75.
Similar paramagnetic behavior was also observed for samples with
x¼0.80, 0.85, 0.90, respectively. The slight asymmetry of the broad
quadrupole splitting doublet is attributed to the contribution of
Fe3þ on the tetrahedral site [39,40].
Progressive substitution of iron by titanium in titanomagnetite
solid solution shows that the ratio of Fe3þ/Fe2þ decreases; the
decrease in the hyperfine magnetic field strength can be
described by the classic substitution between Fe3þ and Ti4þ as
a consequence of Fe3þ reduction:
Fig. 4. Plot of oxidation state of Fe ions in magnetite and ulvo¨spinel Fe2TiO4 as a
function of isomer shift extracted from the corresponding fitted Mo¨ssbauer
spectrum.
2Fe3þ 2Fe2þ þTi4þ
ð2Þ
According to the Mo¨ssbauer spectra, the xFe2TiO4-(1ꢀx)Fe3O4
pseudo-binary solid solutions can be divided into three types
with different magnetic properties: for 0rxr0.4, the solid
solution shows ferrimagnetic behavior, and the Mo¨ssbauer spec-
tra can be de-convoluted as sextets; for 0.4oxr0.7, the solid
solution shows both weak ferrimagnetic and paramagnetic beha-
viors, with dramatic decrease in hyperfine magnetic field strength
of sextets and appearance of quadrupole-split doublet; for
0.7oxr1, the solid solution only shows paramagnetic behavior
and the Mo¨ssbauer spectra consist of quadrupole-split doublets
only. The difference in magnetic properties arises from the degree
of classic substitution between Ti and Fe ions and cation dis-
tribution on tetrahedral/octahedral sites.
For the cation distribution in the xFe2TiO4-(1ꢀx)Fe3O4 solid
solution, it is widely believed that Ti4þ prefers the octahedral site,
which is confirmed by our room temperature Mo¨ssbauer results
that the population percentages of Fe ions on the tetrahedral and
octahedral sites are extremely close to the theoretical value
calculated from the theoretical formula. Though the percentage
of Fe ions on tetrahedral and octahedral site can be determined as
mentioned previously, however, one cannot derive the Fe3þ and
Fe2þ cation distribution in tetrahedral and octahedral sites
directly from the room temperature Mo¨ssbauer spectra due to
the coexistence of Fe3þ and Fe2þ and electron hopping. The area
percentages of de-convoluted sextets and doublets cannot be
used to calculate the Fe3þ and Fe2þ atomic numbers on tetra-
hedral and octahedral sites per solid solution formula, owing to
the strong overlaps between these sub-spectra. However, by the
analysis of isomer shift values of Fe3þ on tetrahedral site in
magnetite, Fe2.5þ on octahedral site in magnetite, and Fe2þ on
either tetrahedral or octahedral sites in the synthesized Fe2TiO4,
respectively, it is very interesting to observe that there is a linear
relationship between the oxidation state of Fe ion and the
corresponding isomer shift, which is shown in Fig. 4.
oxidation state, though it is also affected by the chemical
environment of the atom to some extent [41].
Due to the existence of linear relationship of oxidation state of
Fe ion and the corresponding isomer shift, the average oxidation
state of Fe ions in the synthesized titanomagnetite solid solution
may be derived from the isomer shift of Fe ions in the room
temperature Mo¨ssbauer spectra. This method was also employed
to identify the oxidation states of Fe ions in iron complexes with
cyclam-related ligands [42], a linear relationship between oxida-
tion state of Fe ion and isomer shift was obtained, with oxidation
states of Fe ions in the range of 2þ to 6þ, and the isomer shift
decreasing with the increase in oxidation state. Since the Fe3þ
and Fe2þ ions coexist on either tetrahedral or octahedral site of
the synthesized spinel titanomagnetite, the oxidation state of Fe
ions derived from the linear relationship (Fig. 4) is not an integer,
which in turn reflects the Fe ions distribution. It has been
previously confirmed that Ti4þ prefers the octahedral site for all
of the studied samples and the Fe ion population can be precisely
identified from the fits of room temperature Mo¨ssbauer spectra.
Therefore, we only focus on the isomer shift of Fe ion on the
tetrahedral site. If the Fe cation distribution on tetrahedral site is
obtained, one can derive the formula of the synthesized titano-
magnetite with different molar fraction of ulvo¨spinel components
according to the charge balance. It has to be noted that the effect
of possible defects and impurities in the titanomagnetite samples
on isomer shifts of Fe ions were not considered.
Due to the strong overlap of Mo¨ssbauer sub-spectra it is
extremely difficult to figure out the exact percentage of Fe2þ
and Fe3þ ions from the two de-convoluted sextets or doublets on
tetrahedral site. Instead, the overall isomer shift of Fe ions on the
tetrahedral site was calculated as following, which may reflect
more accurate results about the Fe cation distribution:
Overall isomer shift of Fe ion ¼ ð
d1A1 þ
d2A2Þ=ðA1 þA2Þ
ð3Þ
where stands for the isomer shift and A for area percentage of
d
Isomer shift is a relative measure describing a shift in the
resonance energy of a nucleus due to the transition of electrons
within its s orbital. The whole spectrum is shifted in either a
positive or negative direction depending upon the s electron
charge density. This change arises due to alterations in the
electrostatic response between the non-zero probability s orbital
electrons and the non-zero volume nucleus orbit. Only electrons
in s orbital demonstrate non-zero probability, because their 3D
spherical shape incorporates the volume taken up by the nucleus.
In other words, the s electron density is largely affected by the
corresponding sub-spectra of Fe ions on tetrahedral site. The
calculated overall isomer shifts of Fe ions on the tetrahedral site of
titanomagnetite solid solution are shown in Table 2. According to
the relationship between isomer shift and oxidation state of Fe ions
(Fig. 4), the overall oxidation states of Fe ions on tetrahedral site can
be obtained, and the following rules have to be obeyed with:
3þ
2þ
Overall oxidation state of Fe ions ¼ ðþ3Þ ꢂ Fe
þðþ2Þ ꢂ Fe
apfu
apfu
ð4Þ

M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
1459
Table 2
Overall isomer shift, oxidation state of Fe ion on tetrahedral site, and the derived site occupancy of titanomagnetite with different ulvo¨spinel Fe2TiO4 components.
x
I.S. (mm/s)
Abund.(%)
Overall I.S. (
d
, mm/s)
Overall oxidation state
Derived titanomagnetite formula
(Fe13þ)[Fe12þFe13þ]O4
0.0
0.059
0.472
0.191
0.071
0.318
0.341
0.332
0.373
0.772
0.768
0.904
0.903
70.01
34.6
65.4
27.7
11.1
29.0
11.8
33.4
9.8
0.059
0.472n
0.157
þ3
þ2.5n
2.88
3þ
(Fe Fe2þ )[Fe Fe2þ Ti4þ]O4
0.88 0.12
3þ
0.40
0.60
0.70
0.75
1.0
0.30 1.30 0.4
2þ 3þ
(Fe03.þ68Fe )[Fe Fe2þ Ti4þ]O4
0.32 0.12 1.28 0.6
0.324
0.480
0.768
0.904
70.01
2.68
2.50
2.16
2
2þ
3þ
(Fe03.þ49Fe )[Fe Fe2þ Ti4þ]O4
0.51
0.11 1.19 0.7
2þ
3þ 2þ
8.1
(Fe30.þ005Fe
)[Fe Fe1.25Ti40.þ75]O4
0.995
0.0
38.9
50.3
49.7
70.5
(Fe12þ)[Fe12þTi14þ]O4
Err.
70.02
Note: n 0.472 and þ2.5 are the isomer shift and overall oxidation state of Fe ions on octahedral site in magnetite. The titanomagnetite solid solution with x¼0.65 was not
included due to the large amount of impurities.
distinct exothermic peaks. The first one was assigned to the
oxidation of magnetite to gamma-Fe2O3, while the second corre-
sponds to the change of gamma-Fe2O3 to hematite [43]. Recently,
the Fe2TiO4–Fe3O4 solid solution was studied thermogravimetri-
cally as a function of oxygen activity at high temperature, which
provides information on nonstoichiometry and point defects [44].
However, the thermogravimetric behavior of Fe3O4, Fe2TiO4, and
the synthesized xFe2TiO4-(1ꢀx)Fe3O4 solid solution under inert
atmosphere was not studied, despite its importance to under-
standing the thermal history and magnetic properties of rocks
on Mars.
Similar to the oxidation behavior, the DSC curve of the
magnetite sample shows two distinguished exothermic peaks
and one small endothermic peak (Fig. 6a). The two exothermic
peaks are at 534 1C and 647 1C, respectively. The existence of
exothermic and endothermic peaks on the DSC curve indicates
that magnetite is not stable under heat treatment in inert Ar
atmosphere, suggesting there are either some phase transforma-
tions or crystallization. The first endothermic phase change
occurs below 400 1C with a peak value at 308 1C, while the second
exothermic phase change starts at 400 1C with the peak tempera-
ture at 534 1C. The third exothermic phase transformation takes
place at 570 1C and ends at 700 1C, respectively.
Fig. 5. Cation distribution as a function of molar fraction of ulvo¨spinel component
x in titanomagnetite derived from the relationship between the isomer shift and
oxidation states of Fe ions.
and
As can be seen in Fig. 6h, the TGA curve of magnetite reveals
three major weight losses: one is in the temperature range from
170 1C to 360 1C with a value of ꢁ1.64%, and the other starts from
360 1C and ends at 600 oC, with a weight loss of ꢁ0.70%. The third
weight loss starts at 600 1C and ends at 720 1C, with a value of
ꢁ0.20%. These three weight losses (with the total weight loss
value of ꢁ2.54%) have to be due to the decomposition of
magnetite. The first weight loss is possibly related to the decom-
position of magnetite to non-stoichiometric magnetite, as shown
in Eq. (6). The second and the third weight losses at above 360 1C
are probably due to the further decomposition of non-stoichio-
metric magnetite to another type of non-stoichiometric magne-
tite.
3þ
2þ
Fe
þFe
¼ 1
ð5Þ
apfu
apfu
3þ
Here¼ þ3 and þ2 are the oxidation states of Fe ions, and Fe
and
apfu
2þ
Fe
are the Fe3þ and Fe2þ atomic number per formula unit of
apfu
synthesized titanomagnetite, respectively. The overall isomer shift,
overall oxidation state, and derived titanomagnetite formula are
summarized in Table 2. The plot of Fe cation distributions as a
function of ulvo¨spinel Fe2TiO4 component x is presented in Fig. 5,
which is in good agreement with the reported results from O’Reilly
et al [11]. It has been reported there are different site occupancy
models to describe the Fe2þ and Fe3þ in titanomagnetite solid
solutions [11,14], the difference in those models may arise from
differences in synthesis processes, defects, impurities in titanomag-
netite solid solution and characterization techniques. Room tem-
perature Mo¨ssbauer spectroscopy is another powerful tool to study
the cation distribution in titanomagnetite systems.
Fe3O4-Fe3O4ꢀx þðx=2ÞO2
ð6Þ
These three weight losses are in good agreement with DSC
curves of the original magnetite (Fig. 6a), which shows corre-
sponding endothermic and exothermic peaks.
3.3. Simultaneous DSC–TGA
The DSC–TGA measurement of physically mixed Fe2TiO4–
Fe3O4 powder (with 1:1 weight ratio) was performed for compar-
ison and was shown in Fig. 7. From the TGA curve of this sample,
it was found that the weight loss has a value of ꢁ1.25%, which is
close to half of the weight loss value of original Fe3O4. The weight
loss value for this physically mixed Fe2TiO4–Fe3O4 sample is
The thermal behavior of magnetite on heating has been a
controversial subject, whether magnetite is oxidized to gamma-
Fe2O3 or directly to hematite has caused a lot of discussion. For
magnetite oxidation, it was summarized that DSC curve gives two

1460
M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
Fig. 6. DSC curves of the xFe2TiO4-(1ꢀx)Fe3O4 (0.4rxr1) pseudo-binary system with molar fraction x (a) x¼0.0, (b) x¼0.4, (c) x¼0.6, (d) x¼0.65, (e) x¼0.70, (f) x¼0.75
and (g) x¼1.0, and TGA curves of the xFe2TiO4-(1ꢀx)Fe3O4 (0.4rxr1) pseudo-binary system with molar fraction x (h) x¼0.0, (i) x¼0.4, (j) x¼0.6, (k) x¼0.65, (l) x¼0.70,
(m) x¼0.75 and (n) x¼1.0.
expected by considering the existence of the Fe2TiO4 with a 50%
in weight percentage, and also suggests that the decomposition
behavior of magnetite is not dramatically affected by the Fe2TiO4.
The magnetite will decompose to nonstoichiometric magnetite
under heat treatment and this process is completed below 800 1C.
However, the DSC curve of this physically mixed Fe2TiO4–Fe3O4
sample shows totally different behavior as that of original
magnetite or Fe2TiO4, as no exothermic peak was observed.
Instead, only a broad endothermic peak was observed in the
studied temperature range. The observation of one broad
endothermic peak and non-existence of exothermic peaks sug-
gests the complicated thermal behavior of the physically mixed
Fe2TiO4–Fe3O4 sample. The overall endothermic behavior may be
attributed to the overlap between the decomposition of magne-
tite and the formation of Fe2TiO4–Fe3O4 solid solution under heat
treatment. It also indicates that the magnetite, even with a low
degree of stoichiometry, can also form solid solution with Fe2TiO4
under heat treatment.
that they are thermally stable under heat treatment in Ar atmo-
sphere. The small exothermic peaks on the DSC curves of these
samples may be attributed to the crystallization of fine particles.
The slight variation in the weight percentages on TGA curves is
due to the thermal shift of the weight under heat treatment. The
DSC–TGA results are in good agreement with the XRD results that
no magnetite phase is detected in the xFe2TiO4-(1ꢀx)Fe3O4 solid
solution due to the complete consumption of the original magne-
tite or the non-stoichiometric magnetite.
4. Conclusions
The xFe2TiO4-(1ꢀx)Fe3O4 solid solutions with different molar
fractions x in the range of 0rxr1 were successfully synthesized
by the use of solid-state reaction. X-ray powder diffraction
showed that the synthesized xFe2TiO4-(1ꢀx)Fe3O4 systems main-
tain the inverse spinel structures when Fe ions were replaced by
the Ti4þ. The lattice parameter a expanded linearly with the
increase in the ulvo¨spinel Fe2TiO4 components. 57Fe room
The DSC–TGA curves of the formed xFe2TiO4-(1ꢀx)Fe3O4 solid
solution (with xZ0.4) do not change dramatically, suggesting

M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
1461
Acknowledgments
The work at Duquesne was supported by the National Science
Foundation under grant DMR-0854794.
References
´
[1] L.G. Tejuca, J.L. Fireeo, J.M.D. Tascon, Structure and reactivity of perovskite-
type oxides, Advanced Catalysis 36 (1989) 237–328.
[2] G. Kletetschka, P.J. Wasilewski, P.T. Taylor, Mineralogy of the source for
magnetic anomalies on Mars, Meteoritics and Planetary Science 35 (2000)
895–899.
´
´
[3] A. Wise, M. Saenko, A.M. Velazquez, D.E. Laughlin, M. Dıaz-Michelena,
M.E. McHenry, Phase evolution in the Fe3O4–Fe2TiO4 pseudo-binary system
and its implications for remanent magnetization in Martian minerals, IEEE
Transactions on Magnetics 47 (2011) 4124–4127.
[4] R. Sanz, M. Cerda´n, M.E. McHenry, M. Dı´az-Michelena, Temperature depen-
dent magnetization and remanent magnetization in pseudo-binary
(Fe2TiO4)x–(Fe3O4)1ꢀx (0.30oxo1.00) titanomagnetites, IEEE Transactions
on Magnetics 47 (2011) 4128–4131.
[5] E. Schmidbauer, 57Fe Mo¨ssbauer spectroscopy and magnetization of cation-
deficient Fe2TiO4 and FeCr2O4. Part I: 57Fe Mo¨ssbauer spectroscopy, Physics
and Chemistry of Minerals 14 (1987) 533–541.
[6] E.W. Gorter, Ionic distribution deduced from the g-factor of a ferromagnetic
spinel: Ti4þ in fourfold co-ordination, Nature 173 (1954) 123–124.
[7] V.A. Bokov, A.S. Kamzin, S.I. Yushchuk, N.N. Parfinova, Magnetic fields at 57Fe
nuclei and quadrupole interactions in Ni–Fe–Ti spinels, Soviet Physics —
Solid State 16 (1975) 2360–2363.
[8] M.A. Arillo, M.L. Lo´pez, C. Pico, M.L. Veiga, A. Jime´nez, E. Rodriguez-Castello´ n,
Surface characterization of spinels with Ti(IV) distribution in tetrahedral and
octahedral sites, Journal of Alloys and Compounds 317-318 (2001) 160–163.
[9] S. Akimoto, Thermo-magnetic study of ferromagnetic minerals contained in
igneous rocks, Journal of Geomagnetism and Geoelectricity 6 (1954) 1–14.
[10] L. Ne´el, Some theoretical aspects of rock magnetism, Advanced Physics 4
(1955) 191–242.
[11] W. O’Reilly, S.K. Banerjee, Cation distribution in titanomagnetites (1–x)
Fe3O4–x Fe2TiO4, Physics Letters 17 (1965) 237–238.
[12] Z. Kakol, J. Sabol, J.M. Honig, Cation distribution and magnetic properties of
titanomagnetites Fe3ꢀxTixO4 (0rxo1), Physical Review
B 43 (1991)
649–654.
[13] H.H. Hamdeh, K. Barghout, J.C. Ho, P.M. Shand, L.L. Miller, A Mo¨ssbauer
evaluation of cation distribution in titanomagnetites, Journal of Magnetism
and Magnetic Materials 191 (1999) 72–78.
[14] C.I. Pearce, C.M.B. Henderson, N.D. Telling, R.A.D. Pattrick, J.M. Charnock,
V.S. Coker, E. Arenholz, F. Tuna, G. Van Der Laan, Fe site occupancy in
magnetite-ulvo¨spinel solid solutions: a new approach using X-ray magnetic
circular dichroism, American Mineralogy 95 (2010) 425–439.
[15] N. Millot, S.B. Colin, P. Perriat, G.L. Cae¨r, R. Welter, B. Malaman, Characteriza-
tion of ferrites synthesized by mechanical alloying and soft chemistry,
Nanostructured Materials 12 (1999) 641–644.
[16] D. Basak, J. Ghose, Thermal studies on iron titanate spinel, Journal of Thermal
Analysis 37 (1991) 935–943.
[17] A. Roy, J. Ghose, Studies on thermal stability of titanium substituted iron
molybdenum spinel oxide, Journal. of Thermal Analysis and Calorimetry 61
(2000) 839–847.
Fig. 7. DSC and TGA curves of physically mixed Fe2TiO4–Fe3O4 powder with 1:1
ratio in weight percentage: (a) TGA curve and (b) DSC curve.
temperature Mo¨ssbauer spectra showed that the magnetic prop-
erties of the synthesized solid solutions depend on the molar
fraction of Fe2TiO4 component x and vary significantly, such that
the hyperfine magnetic field decreases with the increase in x
value. When x is in the range of 0rxr0.4, the solid solution
shows ferrimagnetic behavior. However, it shows weak ferrimag-
[18] P. Ramdohr, Ulvo¨spinel and its significance in titaniferous iron ores, Eco-
nomic Geology 48 (1953) 677–688.
netic and paramagnetic behavior with
x in the range of
[19] M.E. Evans, M.L. Wayman, An investigation of the role of ultra-fine titano-
magnetite intergrowth in palaeomagnetism, Geophysical Journal of the Royal
Astronomical Society. 36 (1974) 1–10.
[20] P.P.K. Smith, Spinodal decomposition in a titanomagnetite, American Miner-
alogy 65 (1980) 1038–1043.
[21] L. Ve´gard, Die konstitution der mischkristalle und die raumfu¨llung der atome,
Zeitschrift fu¨r Physik 5 (1921) 17–26.
[22] U. Bleil, An experimental study of the titanomagnetite solid solution series,
Pure and Applied Geophysics 114 (1976) 165–175.
[23] B.C. Melot, K. Page, R. Seshadri, E.M. Stoudenmire, L. Balents, D.L. Bergman,
T. Proffen, Magnetic frustration on the diamond lattice of the A-site magnetic
spinels CoAl2ꢀxGaxO4: the role of lattice expansion and site disorder,
Physical Review B 80 (2009) 104420.
[24] B.A. Wechsler, D.H. Lindsley, C.T. Prewitt, Crystal structure and cation
distribution in titanomagnetite (Fe3ꢀxTixO4), American Mineralogy 69
(1984) 754–770.
0.4oxr0.7. For the solid solution with Fe2TiO4 molar fraction
x40.70, it shows paramagnetic behavior, with appearance of only
quadrupole doublets in Mo¨ssbauer spectra. The accomplishment
and significance of this work are that the Fe cation distribution in
the synthesized xFe2TiO4-(1ꢀx)Fe3O4 systems can be derived
from the isomer shift of room temperature Mo¨ssbauer spectro-
scopy. There is a linear relationship between oxidation state of Fe
ion and isomer shift, in turn, the overall isomer shift value for Fe
ions at tetrahedral sites reflects the overall oxidation state of
Fe ions at tetrahedral site, which can be used to determine the Fe
cation distribution from the demonstrated method.
The simultaneous differential scanning calorimetry and ther-
mogravimetric analysis (DSC–TGA) showed that magnetite is not
stable, and there are some weight losses with exothermic peaks
under heat treatment in inert atmosphere due to the decomposi-
tion of magnetite to non-stoichiometric magnetite. Magnetite
with low degree of stoichiometry can form solid solution with
Fe2TiO4. The synthesized solid solution is stable under heat
treatment in studied temperature range under Ar atmosphere.
[25] F. Bosi, U. Ha˚ lenius, H. Skogby, Crystal chemistry of the magnetite-ulvo¨spinel
series, American Mineralogy 94 (2009) 181–189.
[26] Y. Nakamura, P.A. Smith, D.E. Laughlin, M. De Graef, M.E. McHenry, Structure
and magnetic properties of quenched (MnxAl1ꢀx)3O4 spinels and hausman-
nites, IEEE Transactions on Magnetics 31 (1995) 4154–4156.
[27] Swaminathan, R., McHenry, M.E., Calvin, S., Sorescu, M., Diamandescu,
L.: Surface structure model of cuboctahedrally truncated ferrite nanoparti-
cles. Proc. 9th International Conference on Ferrites, American Ceramic
Society, pp. 847–852, (2005).

1462
M. Sorescu et al. / Journal of Magnetism and Magnetic Materials 324 (2012) 1453–1462
ˇ
ˇ
¨
[28] R. Swaminathan, M.E. McHenry, P. Poddar, H. Srikanth, Magnetic properties
of polydisperse and monodisperse NiZn ferrite nanoparticles interpreted in a
surface structure model, Journal Applied Physics 97 (2005) 10G104.
[29] R. Swaminathan, J. Woods, S. Calvin, J. Huth, M.E. McHenry, Microstructural
evolution model of the sintering behavior and magnetic properties of NiZn
ferrite nanoparticles, Adv. Sci. Tech. 45 (2006) 2337–2344 Online at /http://
www.scientific.netS.
[37] K. Melzer, Z. Simsa, M. Łukasiak, J. Suwalski, Mossbauer spectra and
electronic properties of titanomagnetite, Crystal Research and Technology
22 (1987) K132–K136.
[38] N. Guigue-Millot, Y. Champion, M.J. Hy¨tch, F. Bernard, S. Be´gin-Colin,
P. Perriat, Chemical heterogeneities in nanometric titanomagnetites
prepared by soft chemistry and studied ex situ: evidence for Fe-segregation
and oxidation kinetics, Journal of Physical Chemistry
7125–7132.
B 105 (2001)
[30] E.J.W. Verwey, P.W. Haayman, Electronic conductivity and transition point of
magnetite (Fe3O4), Physica 8 (1941) 979–987.
[39] S.K. Banerjee, W. O’Reilly, T.C. Gibb, N.N. Greenwood, The behaviour of
ferrous ions in iron-titanium spinels, Journal of Physics and Chemistry of
Solids 28 (1967) 1323–1335.
[40] P. Nagarajan, C.K.R. Murty, Mo¨ssbauer and magnetic studies on the system of
Fe3ꢀxTixO4, Bulletin of Materials Science 3 (1981) 217–224.
[41] L.R. Walker, G.K. Werthim, V. Jaccarino, Interpretation of the Fe57 isomer
shift, Physical Review Letters 6 (1961) 98–101.
[31] P.W. Readman, W. O’Reilly, S.K. Banerjee, An explanation of the magnetic
properties of Fe2TiO4, Physics Letters A 25 (1967) 446–447.
[32] M. Tanaka, T. Tokoro, Y. Aiyama, Jahn–Teller effects on Mo¨ssbauer spectra of
Fe57 in FeCr2O4 and FeV2O4, Journal of the Physical Society of Japan 21 (1966)
262–267.
+
´
´
¨
[33] I. Dezsi, I. Szucs, E. Svab, Mossbauer spectroscopy of spinels, Journal of
Radioanalytical and Nuclear Chemistry 246 (2000) 15–19.
[42] J.F. Berry, E. Bill, E. Bothe, S.D. George, B. Mienert, F. Neese, K. Wieghardt,
An octahedral coordination complex of iron (VI), Science 312 (2006)
1937–1941.
[43] K.V.G.K. Gokhale, Study on the oxidation of magnetite, Economic Geology 56
(1961) 963–971.
[34] R. Vanleerberghe, R.E. Vandenberghe, Determination of the quadrupole
splitting distributions of the A and B-site ferrous ions in Fe2TiO4, Hyperfine
Interactions 23 (1985) 75–87.
[35] W. Lowrie, Fundamentals of Geophysics, 1st edition, Cambridge University
Press, Cambridge, UK, 1997 243.
[44] S. Aggarwal, R. Dieckmann, Point defects and cation tracer diffusion in
[36] H. Tanaka, M. Kono, Mo¨ssbauer spectra of titanomagnetite: a reappraisal,
Journal of Geomagnatism and Geoelectricity 39 (1987) 463–475.
(TixFe1ꢀx
)
O4 1. Non-stoichiometry and point defects, Physics and Chem-
3ꢀd
istry of Minerals 29 (2002) 695–706.