Cation distribution in nanocrystalline ZnFe2O4 investigated using x-ray absorption fine structure spectroscopy

ArticleinJournal of Physics Condensed Matter 21(40):405303 · October 2009with133 Reads
Impact Factor: 2.35 · DOI: 10.1088/0953-8984/21/40/405303 · Source: PubMed
Abstract

X-ray absorption fine structure (XAFS) spectroscopy has been employed to investigate the cation distribution in nanocrystalline zinc ferrites (ZnFe(2)O(4)), synthesized in acidic and basic media at different temperatures. By using (Zn(1-x)Fe(x))[Ni(x)Fe(2-x)]O(4) as model compounds we have determined cation distribution in nanosize ZnFe(2)O(4). The cation distribution for samples synthesized at low temperature (400 °C) is (Zn(0.5)Fe(0.5))[Zn(0.5)Fe(1.5)]O(4) for urea- and (Zn(0.75)Fe(0.25))[Zn(0.25)Fe(1.75)]O(4) for citric-acid-based samples. These results show that samples synthesized at and above 600 °C have a local structural environment identical to that of bulk ZnFe(2)O(4).

Full-text

Available from: Muhammad Javed Akhtar
IOP PUBLISHING JOURNAL OF PHYSICS: CONDENSED MATTER
J. Phys.: Condens. Matter 21 (2009) 405303 (9pp) doi:10.1088/0953-8984/21/40/405303
Cation distribution in nanocrystalline
ZnFe
2
O
4
investigated using x-ray
absorption fine structure spectroscopy
M J Akhtar
1,3
, M Nadeem
1
, S Javaid
1
and M Atif
2
1
Physics Division, PINSTECH, PO Nilore, Islamabad, Pakistan
2
Physics Department, Quaid-i-Azam University, Islamabad, Pakistan
E-mail: javeda@pinstech.org.pk
Received 6 April 2009, in final form 19 August 2009
Published 14 September 2009
Online at stacks.iop.org/JPhysCM/21/405303
Abstract
X-ray absorption fine structure (XAFS) spectroscopy has been employed to investigate the
cation distribution in nanocrystalline zinc ferrites
(ZnFe
2
O
4
), synthesized in acidic and basic
media at different temperatures. By using
(Zn
1x
Fe
x
)[Ni
x
Fe
2x
]O
4
as model compounds we
have determined cation distribution in nanosize ZnFe
2
O
4
. The cation distribution for samples
synthesized at low temperature (400
C) is (Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
for urea- and
(Zn
0.75
Fe
0.25
)[Zn
0.25
Fe
1.75
]O
4
for citric-acid-based samples. These results show that samples
synthesized at and above 600
C have a local structural environment identical to that of bulk
ZnFe
2
O
4
.
(Some figures in this article are in colour only in the electronic version)
1. Introduction
The spinel ferrites have attracted intense global interest in
fundamental science such as addressing the relationships
between magnetic properties of nanoparticles and their
crystal structure. The reason for this interest is the
unusual and technologically desirable magnetic properties
of these nanomaterials when compared with their bulk
counterparts [1, 2]. The relatively high magnetization
is explained on the basis of cation distribution between
tetrahedral and octahedral sites in the surface region of the
ultra-fine particles [3]. The synthesis of materials with new
properties by means of the controlled manipulation of their
microstructure on the atomic level has become an emerging
interdisciplinary field based on solid state physics, chemistry,
biology and materials science.
Among spinel ferrites, ZnFe
2
O
4
is of interest not only
to basic research in magnetism, but also has great potential
in technological applications. ZnFe
2
O
4
in bulk form is a
normal spinel, where divalent
(Zn
2+
) ions are at the tetrahedral
(A) sites and trivalent
(Fe
3+
) ions occupy octahedral (B)
sites. However, nanocrystalline ZnFe
2
O
4
shows mixed spinel
3
Author to whom any correspondence should be addressed.
structure, where Zn
2+
and Fe
3+
are distributed over (A) and
(B) sites being represented as
(Zn
2+
1x
Fe
3+
x
)[Zn
2+
x
Fe
3+
2x
]O
4
,
where
x represents the inversion parameter and corresponds
to the degree of cation distribution, i.e., Fe
3+
ions occupying
(A) sites. Figure 1 shows the crystal structure of a typical
spinel ferrite, where tetrahedral (A) and octahedral (B) sites
are shown. It is well known that in bulk ZnFe
2
O
4
the
antiferromagnetic ordering temperature (
T
N
= 10 K) can be
raised by increasing the Fe
3+
occupation at the tetrahedral
sites. This increase of magnetization originates in the
stronger inter-sub-lattice (A–B) superexchange interaction as
compared to intra-sub-lattice (A–A and B–B) interactions. The
ferromagnetic or cluster-glass behaviour of ZnFe
2
O
4
having
a Curie temperature close to 600 K has been reported [4].
Several methods have been employed to synthesize ZnFe
2
O
4
nanoparticles, including ball milling, coprecipitation, aerogel
and hydrothermal methods [5–7]. The magnetic properties
are strongly influenced by the composition and microstructure
of the particles, which are sensitive to the preparation
methodology [8, 9].
The transition temperature
(T
N
) for most ferrites decreases
as the particle size decreases: however, ZnFe
2
O
4
shows
the opposite trend. The origin of this increase has been
variously attributed to oxygen vacancies and disorder [10],
0953-8984/09/405303+09$30.00 © 2009 IOP Publishing Ltd Printed in the UK1
Page 1
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 1. The crystal structure of a typical spinel ferrite (MFe
2
O
4
),
where M is a divalent cation; both tetrahedral (A) and octahedral (B)
sites are shown. In a normal spinal
(ZnFe
2
O
4
) all Zn are at A sites
and all Fe are at B sites, in inverse spinel
(NiFe
2
O
4
) allNiareatB
sites and Fe are equally distributed at A and B sites, whereas in
mixed spinel (nanosized ZnFe
2
O
4
) Zn and Fe are distributed at both
A and B sites, which indicates the degree of inversion.
surface effects [11] and redistribution of Zn or Fe ions at
tetrahedral (A) and octahedral (B) sites [12]. A number
of studies have been carried out to explain the structural
variation due to a decrease in particle size. These
include: x-ray and neutron diffraction [5, 13], M¨ossbauer
spectroscopy [11, 14, 15] and x-ray absorption fine structure
(XAFS) spectroscopy [16–21].
XAFS spectroscopy has revealed itself as a powerful tech-
nique for structural characterization of the local atomic en-
vironment of individual atomic species, including bond dis-
tances, coordination numbers and type of nearest neighbours
surrounding the central atom. This technique is particularly
useful for materials that show considerable structural and
chemical disorder. Previously a number of XAFS studies have
been carried out to investigate the local structural environment
of nanocrystalline ZnFe
2
O
4
. Jeyadevan et al [17] showed that
the local structure of Zn atom is different in bulk and copre-
cipitated ZnFe
2
O
4
; in the latter case, the structure is distorted
due to cation distribution. Oliver et al [18] observed variation
in the local structure around Zn but not around Fe and found
a nonequilibrium distribution of Zn and Fe cations, suggest-
ing overpopulation at octahedral sites. More recently, Stew-
art et al [19] and Nakashima et al [20] employed XAFS spec-
troscopy and used theoretical models to explain the inversion in
nanosized ZnFe
2
O
4
. Although these calculations can simulate
the XANES features which are comparable to experimental
XANES modulations, however, from these calculations precise
XANES spectra with variable degrees of inversion are difficult
to generate. XAFS spectroscopy has been employed to deter-
mine the cation distribution in nanocrystalline CuFe
2
O
4
and
NiFe
2
O
4
systems [22, 23]. The local structural environment of
nanoparticles can be explained on the basis of core and shell,
where the core has bulk characteristics but the structure of the
shell (surface) is different due to variation in the oxygen coor-
dination numbers in the nearest-neighbour shell [24]. There-
fore, in the case of ZnFe
2
O
4
nanostructure, there is a need to
address the surface phenomena as well. In the present study we
use Zn
1x
Ni
x
Fe
2
O
4
as model compounds with known cation
distributions to determine the degree of inversion in nanocrys-
talline ZnFe
2
O
4
. In addition, by using model compounds,
which are polycrystalline, we can make a comparative study
of surface effects on the local environment of nanocrystalline
materials.
It is well known that NiFe
2
O
4
has an inverse spinel
structure with Ni
2+
ions substituted at B sites and Fe
3+
ions
equally distributed at A and B sites. As discussed earlier,
the bulk ZnFe
2
O
4
has a normal spinel structure, with Zn
2+
ions at A sites and all Fe
3+
ions at B sites. Therefore,
when Zn
2+
is substituted with Ni
2+
in Zn
1x
Ni
x
Fe
2
O
4
,mixed
spinel structures having a variable degree of inversion can
be obtained. The cation distribution can be represented
as
(Zn
1x
Fe
x
)[Ni
x
Fe
2x
]O
4
, where ( ) and [ ] are used for
tetrahedral and octahedral sites, respectively. By comparing
the XANES spectra of Zn
1x
Ni
x
Fe
2
O
4
with nanocrystalline
ZnFe
2
O
4
we can accurately determine the degree of inversion.
2. Experimental details
Nanoparticles of ZnFe
2
O
4
were prepared by the sol–gel
method in two different precursors. The chemical reagents
used in the preparation are Zn
(NO
3
)
2
·6H
2
O, Fe(NO
3
)
3
·9H
2
O,
urea and citric acid. All the chemicals are analytical grade.
0.2 M (50 ml) of iron nitrate solution and 0.1 M (50 ml) of
zinc nitrate solution have been used and were gelated by using
0.1 M (300 ml) of urea solution as a catalyst and distilled water
as the solvent. These solutions were heated to a temperature
of 70
C with vigorous stirring until the gel was formed,
which was subsequently dried at 100
Cinanoven. The
nanocrystalline ZnFe
2
O
4
materials were obtained by heating
the dried gel in a muffle furnace at temperatures (400, 600 and
800
C). For the preparation of ZnFe
2
O
4
nanoparticles in an
acidic medium, citric acid was used instead of urea and the
rest of the preparation method was the same as described for
urea-based samples; further details are given elsewhere [8].
The bulk Zn
1x
Ni
x
Fe
2
O
4
(where x = 0.0–1.0) were
synthesized by solid state reaction methods by mixing the
appropriate amount of ZnO, Fe
2
O
3
and NiO (all having
99.99% purity). Acetone was added to the mixture of these
powders, which was finely ground in an electric grinder for
30 min. The powders were heated at 900
C for 15 h in
alumina boats, then cooled to room temperature, reground
and heated at 1100
C for 15 h. The mixture was heated at
1250
C for 15 h twice. The reaction products were furnace-
cooled to room temperature and reground after each heating
cycle. Powder x-ray diffraction (XRD) measurements were
carried out to confirm that single-phase materials have been
prepared.
2
Page 2
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 2. Comparison of the XANES spectra of Fe K-edge in
nanocrystalline ZnFe
2
O
4
and Zn
1x
Ni
x
Fe
2
O
4
. The inset shows zoom
of the pre-edge peak in some selected materials.
The XAFS data for the K-edges of Fe (7112 eV) and
Zn (9569 eV) were collected at beamline 11.1 (XAFS) at the
ELETTRA Synchrotron, Trieste, Italy, with the storage ring
running at 2 GeV and a typical current of 200 mA. These data
were collected at room temperature and in the transmission
mode using an Si(111) monochromator. The Fe and Zn metal
foils were used for energy calibration and to check the stability
of the beamline and optics system. The samples were prepared
by deposition from a powder suspension in cyclohexane on
a Millipore membrane (type GS 0.22
µm). Multi-scanned
spectra of each sample were collected (2–5 scans) to get a
sufficiently high signal-to-noise ratio; these data were merged
together in the first step of the analysis.
The XAFS data were processed by using the Daresbury
data analysis programs EXBACK and EXCURV92 [25, 26].
EXCURV92 employs full curved wave theory [27]and
includes routines to treat multiple scattering effects in highly
symmetric structures. Phase shifts were derived within the
program from ab initio calculations using Hedin–Lundqvist
potentials and von Barth ground states. For each spectrum
a theoretical model was achieved by adding shells around
the central excited atom and least-squares iterating the Fermi
energy
(E
0
), the radial distances (RD) and the Debye–Waller
type factors (DWF). Although the coordination numbers
(CN) can be iterated, due to their strong correlation with
Debye–Waller type factors [28, 29] in the present study
coordination numbers were kept fixed to their preset values.
The structural uncertainties for RD and CN are about
±0.02
˚
Aand±10%, respectively, for nearest neighbours,
increasing slightly for distant shells. The errors in the
fitted parameters are obtained by statistical methods according
to the Error Report of the International XAFS Society
Standards and Criteria Committee 2000 (http://www.i-x-s.org/
OLD/subcommittee
reports/sc/) and the quality of the fit is
measured by an R factor.
3. Results and discussion
In order to compare quantitatively the intensity of absorption
features in nanocrystalline ZnFe
2
O
4
with Zn
1x
Ni
x
Fe
2
O
4
,
used as reference models, the XANES (x-ray absorption near-
edge structure) spectra were normalized. For the normalization
of XANES spectra a standard edge step normalization
procedure was employed [30]. The normalized XANES
spectra were obtained by subtracting the smooth pre-edge
absorption from the experimental spectra and taking the edge
jump height as unity; further details of the normalization
procedure can be found in [30]. The normalized XANES
spectra of the Fe K-edge in nanocrystalline ZnFe
2
O
4
and bulk
Zn
1x
Ni
x
Fe
2
O
4
(x = 0.0–1.0) are shown in figure 2;the
data for Fe foil is also shown, which is used for the energy
calibration. The interpretations of K-edge XANES features
for 3d transition metal oxides are well established [30–33]. A
small pre-edge peak (labelled as A) is observed in all spectra,
which is due to the 1s to 3d (formally electric dipole forbidden)
transition, while the main peak can be attributed to 1s to 4p
transitions [31]. In the case of the Fe K-edge, both 1s to
3d quadrupole transitions and 1s to 4p dipole transitions are
allowed [32], though the intensity of the quadrupole transition
is generally very low. The increase in the intensity of the
pre-edge peak has been attributed to the local mixing of 4p
and 3d orbitals, which is allowed in the tetrahedral symmetry
but forbidden in the octahedral symmetry, due to the presence
of inversion symmetry [31]. This argument is supported by
the DOS calculations where it has been shown that a peak at
2 eV is due to the e
g
level of tetrahedral (A) sites [34]. We
observe that the samples synthesized at 600 and 800
C (both
in urea and citric acid) have identical XANES modulations
and are close to that of bulk ZnFe
2
O
4
. Therefore, we do not
show XANES spectra of samples synthesized in citric acid at
600 and 800
C. From these data we note that the intensity
of the pre-edge peak for 400
C synthesized samples differs
considerably, depending on whether the sample is synthesized
in urea or citric acid. Previous studies have shown that in spinel
ferrites the intensity of the pre-edge peak increases, when there
is an increase in the occupancies of Fe
3+
at the tetrahedral
sites [18, 35]. We use this pre-edge peak to determine the
degree of inversion in nanocrystalline ZnFe
2
O
4
synthesized
under different conditions.
The inset of figure 2 shows the blown-up portion of
the pre-edge peak in nanocrystalline ZnFe
2
O
4
synthesized at
400
C in urea and citric acid along with selected compositions
of the Zn
1x
Ni
x
Fe
2
O
4
series which are used as reference
models having known degrees of inversion. From these results
we observe that the intensity of the pre-edge peak increases
as the concentration of Ni is increased in Zn
1x
Ni
x
Fe
2
O
4
.
This is due to the fact that Ni prefers to go to octahedral (B)
sites; therefore the same amount of Fe moves to tetrahedral
(A) sites. When we compare nanocrystalline ZnFe
2
O
4
with
the bulk Zn
1x
Ni
x
Fe
2
O
4
series we observe that material
synthesized at 400
C in urea has the same pre-edge peak
intensity as that of
(Zn
0.5
Fe
0.5
)[Ni
0.5
Fe
1.5
]O
4
. Therefore the
degree of inversion in this material can be represented as
(Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
. For the sample synthesized at
3
Page 3
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 3. Comparison of the XANES spectra of Zn K-edge in (a) Zn
foil, (b) urea-400, (c) citric-400, (d) urea-600, (e) urea-800 and
(f) bulk ZnFe
2
O
4
.
400
C in citric acid the peak height is comparable to that of
(Zn
0.75
Fe
0.25
)[Ni
0.25
Fe
1.75
]O
4
, indicating that in this material
inversion corresponds to
(Zn
0.75
Fe
0.25
)[Zn
0.25
Fe
1.75
]O
4
.The
peaks for nanocrystalline ZnFe
2
O
4
are relatively broader
compared to the model compounds (which are polycrystalline).
This may be due to the surface effects in nanocrystalline
materials; we will determine it later in the EXAFS discussion.
The intensity of the pre-edge peak for bulk ZnFe
2
O
4
is smaller
than those two compositions, while Zn
0.25
Ni
0.75
Fe
2
O
4
and pure
NiFe
2
O
4
have high pre-edge peak intensities. The present
results clearly demonstrate that, by employing Zn
1x
Ni
x
Fe
2
O
4
as models for mixed spinel structures, we can precisely
determine the degree of inversion in nanocrystalline ZnFe
2
O
4
.
From these results we can infer that ZnFe
2
O
4
nanoparticles
synthesized in urea at 400
C have more inversion compared to
the material synthesized in citric acid at the same temperature.
However, samples synthesized at and above 600
C have a local
structural environment identical to that of bulk ZnFe
2
O
4
and
are independent of the precursor media, acidic or basic.
The normalized XANES spectra of the Zn K-edge in
nanocrystalline ZnFe
2
O
4
prepared at different temperatures
(400, 600 and 800
C) in urea and 400
C in citric acid are
shown in figure 3. As mentioned earlier samples synthesized
above 600
C do not show any variation between urea and
citric acid; therefore, samples prepared in citric acid at 600
and 800
C are not shown here. The data for bulk ZnFe
2
O
4
(synthesized by the solid state method) and Zn foil (which
is used for energy calibration) are also shown. Since Zn
in Zn
1x
Ni
x
Fe
2
O
4
occupies only tetrahedral sites and Ni
alone goes to octahedral sites, therefore the Zn edge of
Zn
1x
Ni
x
Fe
2
O
4
cannot be used as a reference model for
the determination of inversion parameters in nanocrystalline
ZnFe
2
O
4
. In the XANES spectra of the Zn K-edge, there
are three main peaks which are labelled as A, B and C. We
note that peak B is the main peak in the Zn foil and urea-
400 samples and the two shoulder peaks (A and C) are not
Figure 4. EXAFS spectra (a) and Fourier transform (b) of Fe K-edge
in bulk ZnFe
2
O
4
, compared with theoretical model.
yet developed. These peaks start to grow in the citric-400
sample, but the central peak (B) is still dominant. However, for
samples synthesized at 600 and 800
C, these shoulder peaks
become dominant and the central peak is suppressed, which
is comparable to bulk ZnFe
2
O
4
. The present results show
that nanoparticles synthesized at 600
C or above show near-
edge structure identical to that of the bulk material, whereas
for samples synthesized at 400
C, the Zn spectra depend on
whether an acidic or basic medium is used. We note that for the
400
C urea-based sample having (Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
composition, when Zn ions are 50% distributed at both A and B
sites the main central peak is dominant and two shoulder peaks
are not developed. However when the sample is synthesized
in citric acid at 400
C, having (Zn
0.75
Fe
0.25
)[Zn
0.25
Fe
1.75
]O
4
composition, 3/4 Zn ions are at tetrahedral sites and only 1/4
Zn ions are at octahedral sites, which results in the growth of
the shoulder peaks and suppression of the central peak. A peak
(marked as D in spectrum (b) of figure 3)
20 eV above the
edge is prominent in
(Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
(urea 400
C
4
Page 4
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 5. Fe K-edge EXAFS spectra (a) and Fourier transforms (b) in urea-based ZnFe
2
O
4
, along with bulk ZnFe
2
O
4
and Zn
0.5
Ni
0.5
Fe
2
O
4
.
synthesized sample), which is identical to what was observed
in ZnO and has been attributed to a phase transition [36].
A similar peak has been observed in the simulated spectrum
when an isolated Zn atom is substituted for Fe at B sites in
ZnFe
2
O
4
[18]. However, we note that the simulated peak
position is 3 eV higher than our experimental value. The
difference in the peak positions could be due to the fact that
in simulation an isolated Zn atom is substituted at Fe sites.
From our experimental data we observe that the intensity of
this peak gradually decreases as the concentration of Zn at
(B) sites decreases; therefore we can infer that this peak has
contributions from Zn
2+
ions occupying the octahedral sites in
nanocrystalline ZnFe
2
O
4
.
EXAFS spectra and Fourier transforms (both experimental
and fitted models) of the Fe K-edge in ZnFe
2
O
4
are shown
in figure 4 (to compare with the theoretical model the data
are Fourier-filtered). The results obtained from the best fitted
model are presented in table 1. These results are in good
agreement with the bond lengths obtained from the lattice
Table 1. Radial distribution of Fe K-edge in ZnFe
2
O
4
.
Bulk ZnFe
2
O
4
Urea-800
Shell CN RD (
˚
A) DWF (
˚
A
2
) RD (
˚
A) DWF (
˚
A
2
)
Fe–O 6.0 2.02 ± 10.010 ± 12.01 ± 10.011 ±1
Fe–Fe 6.0 2
.99 ± 10.010 ± 12.99 ± 10.011 ±1
Fe–O 2.0 3
.47 ± 20.004 ± 33.48 ± 20.004 ±3
Fe–Zn 6.0 3
.51 ± 30.019 ± 43.52 ± 10.020 ±3
Fe–O 6.0 3
.61 ± 20.018 ± 93.60 ± 30.038 ±7
R factor =10.11% R factor = 10.23%
parameters [37]. For the Fe edge, we note that there is one
peak at
2.0
˚
A which is due to the first shell of oxygen
atoms, the second peak at
3.0
˚
A is due to Fe second-
nearest neighbours, while the third peak at
3.5
˚
A is due to
the combination of Zn and oxygen atoms. The theoretical
model presented in table 1 can be fitted to nanocrystalline
ZnFe
2
O
4
synthesized at 600 and 800
C (only the results
for samples synthesized at 800
C are presented). The
5
Page 5
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 6. Fe K-edge EXAFS spectra (a) and Fourier transforms (b) in citric-acid-based ZnFe
2
O
4
, along with bulk ZnFe
2
O
4
and
Zn
0.75
Ni
0.25
Fe
2
O
4
.
EXAFS spectra and Fourier transforms of the Fe K-edge
for ZnFe
2
O
4
(bulk and nanoparticles prepared in urea at
different temperatures) along with Zn
0.5
Ni
0.5
Fe
2
O
4
are shown
in figure 5. We note that in the case of the urea-400 sample
the EXAFS amplitude reduces significantly at higher k space
(figure 5(a)). Since in EXAFS the amplitude is proportional
to the coordination number, therefore these results show that
coordination numbers reduce in the case of nanocrystalline
ZnFe
2
O
4
, suggesting loss of long range order. When we
look at the Fourier transforms it is evident that the intensity
of the first peak is reduced in 400
C synthesized samples;
in addition, the second and third peaks are not resolved.
We observe that both peaks have low peak intensities in the
urea 400
C sample, (Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
(figure 5(b)).
The intensity of the first peak (which is due to first-shell
oxygen atoms) is comparable with that of the model compound
(Zn
0.5
Ni
0.5
Fe
2
O
4
), clearly indicating that the degree of
inversion is the same in both samples. However, the EXAFS
oscillation in Zn
0.5
Ni
0.5
Fe
2
O
4
is similar to bulk material even
at higher
k space. From these results we note that in the
case of the urea-400 sample the surface effects are significant
and in addition to inversion there is no long range order
present in this material. These effects are also observed in the
Fourier-transform spectra where the magnitude of the Fourier
transform is reduced considerably after 4
.0
˚
A for the urea-400
sample.
Figure 6 shows the comparison of EXAFS spectra and
Fourier transforms of nanocrystalline ZnFe
2
O
4
synthesized in
citric acid along with bulk ZnFe
2
O
4
and Zn
0.75
Ni
0.25
Fe
2
O
4
.
Here we again note that the intensity of the first peak
is reduced in the citric-400 sample compared to the other
samples, although the EXAFS oscillations (figure 6(a)) are
low, but not as much as was observed in the urea-400
sample. From the Fourier transforms it is observed that
the intensity of the first peak in the citric-400 sample is
reduced (figure 6(b)) when compared with bulk ZnFe
2
O
4
and
high temperature synthesized samples. When these peaks
are compared with Zn
0.75
Ni
0.25
Fe
2
O
4
, we note there is good
6
Page 6
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 7. EXAFS spectra (a) and Fourier transform (b) of Zn K-edge
in bulk ZnFe
2
O
4
, compared with theoretical model.
agreement between the two peaks. From these results we
can deduce that the sample synthesized in citric acid at
400
C has inversion of (Zn
0.75
Fe
0.25
)[Zn
0.25
Fe
1.75
]O
4
.For
the citric-400 sample we also note that the magnitude of the
Fourier transform is relatively low after 4
.0
˚
A compared to
the Zn
0.75
Ni
0.25
Fe
2
O
4
which can be inferred from the surface
effects in nanocrystalline materials.
The results of Fe K-edge EXAFS data analysis show
that there is no considerable change in the local environment
of nanocrystalline materials when synthesized at and above
600
C, compared with bulk ZnFe
2
O
4
. However, from these
results we observe that when ZnFe
2
O
4
is synthesized at
low temperature (400
C) there is a local structural change
around Fe, which arises due to different degrees of inversion
and surface effects in nanocrystalline ZnFe
2
O
4
. It is clear
from these findings that ZnFe
2
O
4
nanoparticles having various
degrees of inversion can be obtained by applying different
synthesis conditions.
Table 2. Radial distribution of Zn K-edge in ZnFe
2
O
4
.
Bulk ZnFe
2
O
4
Urea-800
Shell CN RD (
˚
A) DWF (
˚
A
2
) RD (
˚
A) DWF (
˚
A
2
)
Zn–O 4.0 1.98 ± 10.009 ± 11.98 ± 10.009 ± 1
Zn–Fe 12.0 3
.51 ±20.015 ± 23.52 ± 10.014 ± 2
Zn–O 12.0 3
.53 ±30.031 ± 43.53 ± 20.026 ± 5
Zn–Zn 4.0 3
.65 ±20.014 ± 23.66 ± 20.014 ± 2
R factor = 22.59% R factor =23.52%
In figure 7 EXAFS spectra and Fourier transforms (both
experimental and fitted model) of the Zn K-edge in bulk
ZnFe
2
O
4
are shown (to compare with the theoretical model the
data are Fourier-filtered). The results obtained from the best fit
are presented in table 2: for comparison, data obtained from
nanoparticles synthesized in urea at 800
Cisalsogiven. For
nanocrystalline materials synthesized at 600 and 800
Cthis
model can be fitted; we do not observe any appreciable change
between the local structural environments of nanoparticles and
bulk ZnFe
2
O
4
. The EXAFS spectra and Fourier transforms of
Zn K-edge in nanocrystalline ZnFe
2
O
4
, prepared at different
temperatures, in urea and citric acid along with the bulk sample
are shown in figure 8. We note that the amplitude, frequency
and phase of the EXAFS spectrum of the urea 400
C
synthesized sample are different from Zn
0.5
Ni
0.5
Fe
2
O
4
and
bulk ZnFe
2
O
4
spectra (figure 8(a)). This is due to the fact that
in Zn
0.5
Ni
0.5
Fe
2
O
4
all Zn are at A sites (as in the case of bulk
ZnFe
2
O
4
), whereas in the urea-400 sample Zn are distributed
at both A and B sites having
(Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
composition. The phase and amplitude of the EXAFS provide
information about the type of scattering atoms [38]; which
are different when Zn is at a tetrahedral site or occupies an
octahedral site, having inversion symmetry. This clearly shows
that the Zn K-edge in Zn
0.5
Ni
0.5
Fe
2
O
4
cannot be used as
a model for the determination of local environment around
Zn in ZnFe
2
O
4
nanocrystalline materials. From Fourier
transforms (figures 8(b) and (c)) we observe that for all
samples there is not much difference in the intensity of the
first peak. However, we note that for samples synthesized
at 400
C, the amplitude of the second peak decreases
considerably; furthermore, in the case of the urea-400 sample,
(Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
, there is a shift in the peak position
towards lower radial distance. This can be expected, because
when Zn moves from tetrahedral to octahedral sites the second-
nearest neighbour distance decreases
3.5to3.0
˚
Aand
the number of second-nearest neighbours reduces from 12
to 6. For materials synthesized at 400
C in citric acid,
(Zn
0.75
Fe
0.25
)[Zn
0.25
Fe
1.75
]O
4
, the structural changes are less
prominent, because 3
/4 Zn atoms are at tetrahedral sites
and only 1
/4 Zn atoms occupy the octahedral symmetry.
From these results we can infer that contributions from
second-nearest neighbours vary in nanocrystalline ZnFe
2
O
4
,
where tetrahedral symmetry is considerably different from the
octahedral symmetry.
7
Page 7
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
Figure 8. Zn K-edge EXAFS spectra (a) and Fourier transforms (b) of urea-based ZnFe
2
O
4
, along with bulk ZnFe
2
O
4
and Zn
0.5
Ni
0.5
Fe
2
O
4
and Fourier transforms of citric acid based ZnFe
2
O
4
, along with bulk ZnFe
2
O
4
and Zn
0.75
Ni
0.25
Fe
2
O
4
(c).
4. Conclusion
From the above results we can conclude that by employing
XAFS spectroscopy we can precisely determine the cation
distribution in nanocrystalline ZnFe
2
O
4
,usingZn
1x
Ni
x
Fe
2
O
4
as model compounds. We found that for samples synthesized
at low temperature (400
C) the degree of inversion depends
on the precursor medium. For urea-based samples the
degree of inversion is found to be
(Zn
0.5
Fe
0.5
)[Zn
0.5
Fe
1.5
]O
4
,
whereas for materials synthesized at 400
C in citric acid, the
cation distribution is
(Zn
0.75
Fe
0.25
)[Zn
0.25
Fe
1.75
]O
4
.However,
materials synthesized at 600
C and above have a local
structural environment identical to that of bulk ZnFe
2
O
4
.
These results clearly demonstrate that the use of model
compounds with known cation distribution can be employed
to determine the degree of inversion in nanocrystalline spinel
ferrites. In addition, these results show that, apart from
inversion, surface effects are prominent in nanocrystalline
ZnFe
2
O
4
synthesized in urea at 400
C.
Acknowledgments
We acknowledge the support of Elettra Synchrotrone, Trieste,
Italy, for the provision of beam time and the ICTP-Elettra
users’ programme for a financial grant. We are grateful to
Dr Luca Olivi for his valuable suggestions and help during
XAFS data collection.
References
[1] Gleiter H 2000 Acta Mater. 48 1
[2] Moriarty P 2001 Rep. Prog. Phys. 64 297
[3] van der Zaag P J, Brabers V A M, Johnson M T,
Noordermeer A and Bongers P F 1995 Phys. Rev. B
51 12009
[4] Chen Y F, Spoddig D and Ziese M 2008 J. Phys. D: Appl.
Phys. 41 205004
[5] Kamiyama T, Haneda K, Sato T, Ikeda S and Asano H 1992
Solid State Commun. 81 563
[6] Ho J C, Hamdeh H H, Chen Y Y, Lin S H, Yao Y D,
Willey R J and Oliver S A 1995 Phys. Rev. B 52 10122
[7] Shenoy S D, Joy P A and Anantharaman M R 2004 J. Magn.
Magn. Mater. 269 217
[8] Atif M, Hasanain S K and Nadeem M 2006 Solid State
Commun. 138 416
[9] Philip J, Gnanaprakash G, Panneerselvam G, Antony M P,
Jayakumar T and Raj B 2007 J. Appl. Phys. 102 054305
[10] Goya G F and Rechenberg H R 1999 J. Magn. Magn. Mater.
196/197 191
[11] Li F S, Wang L, Wang J B, Zhou Q G, Zhou X Z, Kunkel H P
and Williams G 2004 J. Magn. Magn. Mater. 268 332
[12] Jayadevan B, Tohji K, Nakatsuka K and Narayanasamy A 2000
J. Magn. Magn. Mater. 217 99
[13] Shimada T, Tachibana T, Nakagawa T and Yamamoto T A
2004 J. Alloys Compounds 379 122
8
Page 8
J. Phys.: Condens. Matter 21 (2009) 405303 M J Akhtar et al
[14] Upadhyay C, Verma H C and Anand S 2004 J. Appl. Phys.
95 5746
[15] Oliver S A, Hamdeh H H and Ho J C 1999 Phys. Rev. B
60 3400
[16] Calvin S, Carpenter E E, Ravel B, Harris V G and Morrison S A
2002 Phys. Rev. B 66 224405
[17] Jeyadevan B, Tohji K and Nakatsuka K 1994 J. Appl. Phys.
76 6325
[18] Oliver S A, Harris V G, Hamdeh H H and Ho J C 2000 Appl.
Phys. Lett. 76 2761
[19] Stewart S J, Figueroa S J A, Ramallo L´opez J M, Marchetti S
G, Bengoa J F, Prado R J and Requejo F G 2007 Phys. Rev.
B 75 073408
[20] Nakashima S, Fujita K, Tanaka K, Hirao K, Yamamoto T and
Tanaka I 2007 Phys. Rev. B 75 174443
[21] Henderson C M, Charnock J M and Plant D A 2007 J. Phys.:
Condens. Matter 19 076214
[22] Krishnan V, Selvan R K, Augustin C O, Gedanken A and
Bertagnolli H 2007 J. Phys. Chem. C 111 16724
[23] Carta D, Loche D, Mountjoy G, Navarra G and Corrias A 2008
J. Phys. Chem. C 112 15623
[24] Wu Z, Zhang J, Chen X, Chen Z, Sun M, Wu Z and
Guo L 2005 Phys. Scr. T 115 802
[25] Morrel C, Baines J T M, Campbell J C, Diakun G P,
Dobson B R, Greaves G N and Hasanain S S 1988 SERC
Daresbury Laboratory, EXAFS Users Manual
[26] Stephenson P, Binsted N, Gurman S J and Campbell J W 1992
SERC Daresbury Laboratory, EXCURV92 Program
[27] Gurman S J, Binsted N and Ross I 1984 J. Phys. C: Solid State
Phys. 17 143
[28] Akhtar Z N, Akhtar M J and Catlow C R A 1994 J. Mater.
Chem. 4 1081
[29] Akhtar M J, Shaheen R, Haque M N, Bashir J and Akhter J I
2000 Supercond. Sci. Technol. 13 1612
[30] Wong J, Lyte F W, Messmer R P and Maylotte D H 1984 Phys.
Rev. B 30 5596
[31] Groot de F 2001 Chem. Rev. 101 1779
[32] Westre T E, Kennepohl P, DeWitt J G, Hedman B,
Hodgson K O and Solomon E I 1997 J. Am. Chem. Soc.
119 6297
[33] Chaurand P, Rose J, Briois V, Salome M, Proux O, Nassif V,
Olivi L, Susini J, Hazemann J L and Bottero J Y 2007
J. Phys. Chem. B 111 5101
[34] Anisimov V I, Elfimov I S, Hamada N and Terakura K 1996
Phys. Rev. B 54 4387
[35] Bonsdorf G, Denecke M A, Schafer K, Christen S,
Langbein H and Gunber W 1997 Solid State Ion. 101–103
351
[36] Liu H, Yang J, Zhang Y, Yang L, Wei M and Ding X 2009
J. Phys.: Condens. Matter 21 145803
[37] Schiessl W, Potzel W, Karzel H, Steiner M, Kalvius G M,
Martin A, Krause M K, Halevy I, Gal J, Schafer W, Will G,
Hillberg M and Wappling R 1996 Phys. Rev. B 53 9143
[38] Penner-Hahn J E 1999 Coord. Chem. Rev. 190–192 1101
9
Page 9
    • "The magnetic properties of ferrites are sensitive to their composition, processing techniques, microstructure and cations distribution within the crystal structure [5]. In general there is not a complete inversion of cations between tetrahedral and octahedral sites in the spinel structure, but a partial inversion which could be written as follows (Me 1À x Fe x ).(Fe 2 À x Me x )O 4 [6,7]. Theoretical simulations suggest that the electrostatic energy of the system plays a crucial role in deciding the cationic configuration of spinel ferrites [7]. "
    [Show abstract] [Hide abstract] ABSTRACT: In present report, we have discussed the effect of Cr 3 þ substitution on structural and magnetic properties of ZnFe 2 À x Cr x O 4 (x ¼ 0, 0.1, 0.3, 0.5) ferrite nano-particles synthesized by a chemical co-precipitation method using oleic acid as surfactant. Structural studies were carried out using X-ray diffraction (XRD) and transmission electron microscopy (TEM) techniques. XRD pattern of the prepared powder samples annealed at 350 1C confirmed the formation of single phase cubic spinel structure. TEM images confirmed the nano size formation of particles. A Vibrating Sample Magnetometer (VSM) has been employed to probe the change in magnetic properties with change in composition. Magnetic particle size (D m) calculated from hysteresis loop was found less than the particle size (D T) calculated from TEM micrograph due to spin canting effect. The magnetization studies inferred that magnetic moment (η B) showed decreasing trend with increase in Cr 3 þ substitution, its value decreases from 0.548 (for x ¼ 0.0) to 0.125 (for x ¼0.5). The saturation magnetization (M S) value was found to decrease from 12.76 emu/gm (for x ¼0.0) to 2.92 emu/gm (for x ¼0.5). The coercivity (H c) was found to change proportionally with the particle sizes of investigated ferrites. Squareness (S) values revealed that particles interact by magnetostatic interactions. The M–H loop of all the samples is narrow with low value of coercivity and retentivity; indicate the superparamagnetic nature of these samples. &
    Full-text · Article · Jan 2015 · Ceramics International
    • ". The inversion parameter 'd', is characterized as the fraction of A-sites occupied by Fe 3? cations and its value depends on the methods of preparation [5, 6]. Sol–gel method is a versatile technique to vary the properties of the material by controlling different parameters such as temperature, time of reaction, pH of the medium and the reagent's concentration [7]. "
    [Show abstract] [Hide abstract] ABSTRACT: Zn1-xNixFe2O4 (0.0 B x B 1.0) nanoparticles are prepared by sol–gel method using urea as a neutralizing agent. The evaluation of XRD patterns and TEM images indicated fine particle nature. The average crystallite size increased from 10 to 24 nm, whereas lattice parameters and density decreased with increasing Ni content (x). Infrared spectra showed characteristic features of spinel structure along with a strong influence of compositional variation. Magnetic measurements reveal a maximum saturation magnetization for Zn0.5Ni0.5Fe2O4 (x = 0.5); however, reduced value of magnetization is attributed to the canted spin structure and weakening of Fe3?(A)–Fe3?(B) interactions at the surface of the nanoparticles. Impedance analysis for different electro-active regions are carried out at room temperature with Ni substitution. The existence of different relaxations associated with grain, grain boundaries and electrode effects are discussed with composition. It is suggested that x = 0.5 is an optimal composition in Zn1-xNixFe2O4 system with moderate magnetization, colossal resistivity and high value of dielectric constant at low frequency for their possible usage in field sensor applications.
    Full-text · Article · Aug 2014 · Journal of Sol-Gel Science and Technology
    • "In the present study the Fe K-edge data has been used to investigate the valence state of Fe in SrFe 0.5 Nb 0.5 O 3 and SrFeO 3 ; we have taken Fe 2 O 3 and Fe 3 O 4 as model compounds. In Fe 2 O 3 the valence state of Fe is III whereas in Fe 3 O 4 it is a mixture of II and III.Fig. 4 shows the normalized XANES spectra of these materials along with Fe metal foil, which is used for the energy calibration [24,27]; from these results we note that there is a pronounced shift in the edge position. In case of Fe 3 O 4 spectrum, edge is about ~1 eV below Fe 2 O 3 and SrFe 0.5 Nb 0.5- O 3 , both these spectra have same energy shift at half of the normalized edge height. "
    [Show abstract] [Hide abstract] ABSTRACT: The perovskite based SrFeO{sub 3} and SrFe{sub 0.5}Nb{sub 0.5}O{sub 3} materials have been synthesized by solid state reaction methods. The structural properties are investigated using a combination of X-ray diffraction and X-ray absorption fine structure spectroscopic techniques. From the Rietveld refinement of the X-ray diffraction data it has been observed that SrFeO{sub 3} has a simple cubic perovskite structure, which is consistent with the previous literature results; whereas SrFe{sub 0.5}Nb{sub 0.5}O{sub 3} shows a tetragonal structure within P4mm space group. X-ray absorption results demonstrate that the valence state of Fe in SrFeO{sub 3} is (IV); however, it changes to (III) when 50% Nb{sup 5+} is substituted at the Fe sites. - Highlights: {yields} Structural studies by employing XRD and XANES spectroscopic techniques. {yields} Rietveld refinement confirmed SrFeO{sub 3} has cubic structure, space group Pm-3m. {yields} It is revealed that SrFe{sub 0.5}Nb{sub 0.5}O{sub 3} has tetragonal structure, in P4mm space group. {yields} From XANES results it is observed that Fe has valence state of (IV) in SrFeO{sub 3}. {yields} Doping of 50% Nb{sup 5+} at Fe sites, changes Fe valence to (III) in SrFe{sub 0.5}Nb{sub 0.5}O{sub 3}.
    Full-text · Article · Oct 2011 · Materials Characterization
Show more

Similar publications

Discover more