ARTICLE IN PRESS
634
T.P. Braga et al. / Journal of Magnetism and Magnetic Materials 322 (2010) 633–637
derived from
D-glucosamine, as an organic precursor, and
aluminum and iron salts: 5.5 g of chitosan was dissolved in
300 mL of CH3COOH solution (5% v/v); separately, 26.98 g of
Al(NO3)3 Á 9H2O and 4.85 g of Fe(NO3)3 Á 9H2O were dissolved in
100 mL of water. The iron + aluminum aqueous solution was
poured into the chitosan solution, under constant stirring. The
resulting solution, from now on to be referred to as simply Fe–Al–
chitosan solution, was drop wise pumped into a NH4OH aqueous
solution (30% v/v), under stirring, with a peristaltic pump. The gel
spheres so formed were separated from the NH4OH solution
medium and dried at room temperature for 96 h. The dried
samples were calcined in airflow, during 1 h at 500 1C, under a
heating rate of 5 1C minÀ1. During the polymeric precursor
elimination process, the iron and/or aluminum oxide spheres
were formed. The Fe–Al–chitosan solutions were prepared with a
ratio of 2.5 ions (Fe and Al) to each monomer of chitosan. Samples
with different Al to Fe molar ratios were prepared and labeled
AlFeX, where X denotes the Al:Fe molar ratio. The sample labeled
Al contained only aluminum oxide.
Fig. 1. Spheres immediately after the preparation process (AlFe6).
are rather small. Two crystalline phases of iron oxide were
identified from the diffraction profile of sample AlFe0 (see Fig.
aFe2O3 (hematite, JCPDS card # 87-1166) and gFe2O3
(maghemite, JCPDS card # 25-1402). Table 1 shows proportions
of the main occurring phases.
The morphology and mean diameter of spheres were exam-
ined with a Philips XL30 scanning electron microscope (SEM),
operating with an accelerating voltage of 20 kV. The X-ray
diffraction (XRD) analysis was performed in a Rigaku-DMAXB X-
3b):
ray diffractometer using Bragg–Brentano geometry in the range of
˚
The diffraction patterns for samples AlFe15, AlFe6, and Al (Fig.
3a) show the occurrence of Al2O3 (JCPDS card # 10-0425); figures
for sample AlFe6 and AlFe2 reveal also the co-existence of iron
10–801 with a rate of 0.5 1minÀ1. CuK
a radiation (l=1.5405 A)
was used and the tube operated at 40 kV and 25 mA. The phase
identification analysis was made by comparing obtained powder
diffractograms with standard patterns from International Centre
for Diffraction Data (ICDD). For the sample AlFe0, the experi-
mental patterns were numerically fitted with the Rietveld
algorithm [20] in a procedure to better identify and quantify
crystallographic phases. Mean nanoparticles sizes, when applied,
were estimated by using the Scherrer’s equation [21]. Transmis-
sion Mo¨ssbauer spectra were recorded at room temperature,
110 K, and at 20 K, in constant acceleration mode setup, with a
57Co (Rh) source. The Mo¨ssbauer data were fitted to discrete
Lorentzian functions, using the least-square fitting routine of the
oxide phases, namely
aFe2O3 (hematite, JCPDS card # 87-1166)
and Fe2O3 (maghemite, JCPDS card # 25-1402). Reflections
g
associated with Al-bearing phases were not readily identified
from the pattern for sample AlFe2, whereas Fe-bearing phases
were not identified in the pattern for sample AlFe15. This is not
unexpected as samples AlFe2 and AlFe15 have, respectively, the
highest and lowest Fe:Al ratios of all samples containing
simultaneously the two elements. These results suggest that the
high aluminum ratio inhibits the formation of iron-containing
compounds, such as hematite or magnetite, and are consistent
with other reportedly results [22,23]. This influence may be
NORMOSs software package. All isomer shift values (
quoted relatively to Fe.
d) are
explained in terms of the ionic radii of the elements as the radius
3+
of octahedral Al3+ (0.53 A) is comparable to that of Fe (0.67 A).
Similar ionic radii favor the insertion of isomorphic Al3+ into the
structure of iron oxide, but this replacement tends to inhibit the
hematite crystallization.
˚
˚
a
3. Results and discussion
Any attempt to perform the Rietveld refinement of XRD data
for samples Al, AlFe15, AlFe6, and AlFe2 did not yield reliable
results as samples are poorly crystalline. The average crystallite
diameter of the different phases as estimated with the Scherrer’s
formula [21] is shown in Table 1. The mean coherent lengths
(MCL) for Al2O3 in samples Al, AlFe15, and AlFe6 were found to be,
Fig. 1 illustrates the hybrid spheres immediately after being
separated from the aqueous solution of ammonia. The average
diameter of spheres, determined at this stage, is 3.08 mm with
standard deviation of 0.34; this mean value was determined
considering a total of 350 spheres. Therefore, a reasonably
uniformity of sizes throughout the entire sample mass is
confirmed from the image (Fig. 1). After the drying process, the
mean diameter of spheres is significantly reduced, as it can also be
observed from SEM micrographs (Fig. 2), indicating that volumes
of individual sphere are significantly influenced by water retained
by particles during the chemical synthesis. After drying at room
temperature, the mean diameters for samples AlFe6 and AlFe0
were found to be 1.42 and 1.69 mm, respectively. After calcination
at 5001C under airflow, the mean diameters for samples AlFe6 and
AlFe0 became 1.02 and 1.53 mm, respectively (Fig. 2a and c). The
occurrence of some particles surface cracked (Fig. 2b and d) points
to the need of improving further the mechanical resistance of the
material. However, this is an issue being addressed in a future
report, as corresponding data are still being more accurately
collected.
respectively, 2.0, 2.0, and 2.9 nm; for
AlFe2, 2.8 and 3.0 nm in diameter; for
MCL values for Fe2O3 and Fe2O3 in the more crystalline AlFe0
a
Fe2O3 in samples AlFe6 and
gFe2O3, 2.6 and 2.5 nm.
a
g
were found as being 9.9 and 9.0 nm. Data in Table 1 suggest that
the increasing content of aluminum oxide tends to decrease in the
diameter of iron oxide crystallites.
The local environment of iron atoms in the iron-containing
samples was investigated by Mo¨ssbauer spectroscopy. Fig. 4
shows the fitted spectra; the corresponding hyperfine parameters
are presented in Table 2. The room temperature spectra obtained
for samples AlFe15, AlFe6, and AlFe2 (Fig. 4a) show similar
features with a central doublet suggesting that iron is in a (super)
paramagnetic state. The Mo¨ssbauer signal for sample AlFe15
corroborates the assumption that Fe-based phases are not
identifiable by XRD due to low Fe:Al ratio and to the extremely
small particle sizes.
The calcinated samples were also analyzed with X-ray
diffraction; results are presented in Fig. 3. All patterns, except
for sample AlFe0, present broad peaks, indicating that crystallites
XRD analysis confirms the existence of very fine-grained
aFe2O3 and gFe2O3 (Table 1) in all samples, whereas the doublets
Braga, Tiago P.
