Full text of DOI:10.1557/JMR.2000.0101

Preparation of nanocomposites containing iron and
nickel–zinc ferrite
M. Pal
Indian Association for the Cultivation of Science, Jadavpur, Calcutta, 700 032, India
D. Das and S.N. Chintalapudi
Inter University Consortium for DAE facilities, LB8 Sector 3, Bidhan Nagar, Calcutta, 700 031, India
a)
D. Chakravorty
Indian Association for the Cultivation of Science, Jadavpur, Calcutta, 700 032, India
(Received 18 March 1999; accepted 6 December 1999)
Composites consisting of nanometer-sized nickel–zinc ferrite and ␣-iron were prepared
by subjecting micrometer-sized ferrite particles to a reduction treatment in the presence
of ␣–Fe O . The materials were characterized by x-ray diffraction, electron microscopy,
2
3
Mossbauer spectroscopy, and magnetization measurements. A wide range of saturation
magnetization and coercivity can be obtained by changing the reduction schedule. The
reduction process appears to break down the particle size of the precursor powder of
nickel–zinc ferrite.
I. INTRODUCTION
the ferrite phase being known to exhibit a semiconduct-
ing behavior due to the presence of different valence
Composites consisting of nanometer-sized crystals in a
host medium have been studied extensively in recent
1
0
states of iron. It is expected that these composites
should also exhibit interesting electrical properties mak-
ing them suitable for applications in electronic systems
as well. However, in this paper we concentrate only on
the magnetic behavior of these materials.
1–4
years. Iron particles having sizes in the range 2–10 nm
and dispersed in an insulating matrix such as SiO or
2
5–7
Al O exhibit giant magnetic coercivity.
Nanocom-
2
3
posite films consisting of ferromagnetic nanoparticles
embedded in an antiferromagnetic oxide have been
previously investigated because of their potential appli-
cation as magnetic recording media, due to useful char-
acteristics such as high coercivity, small grain size, and
II. EXPERIMENTAL
First, a powder with the target composition 0.4 NiO,
0
.4 ZnO, and 1.2 Fe O was prepared. The resultant
2 3
8
resistance to corrosion. These materials were deposited
material after heat treatment would have an Fe O con-
2
3
by direct current (dc) magnetron sputtering of a metal
phase (cobalt) on suitable substrates. The metal parti-
cles were then oxidized to form a surface layer of an-
tiferromagnetic cobalt oxide. We had previously grown
nanocrystalline metal species at the surfaces of oxide
crystalline phases within a glass-ceramic. We have now
investigated the possibility of preparation of composites
consisting of a nanocrystalline ferromagnetic phase like
tent higher than what is required in nickel–zinc ferrite;
that is, (Ni0.5 Zn )Fe O . It should be noted that the
0.5
2 4
desired ferrite was of the latter composition. Though
the initial strategy was to keep a higher Fe O content
2
3
than necessary for ferrite formation for obtaining ␣–Fe
after reduction treatment, our subsequent experiments
showed that this was not at all necessary. The results
in terms of the different phase formation as reported
in the following sections remained unchanged even if
no excess amount of Fe O was taken for sample prepa-
9
␣–Fe and a ferrimagnetic oxide phase like nickel–zinc
ferrite. The method adopted here is that of a reduction
treatment applied to the oxide phase. Essentially, a con-
ventional ceramic processing route has been adopted
here. It appears that a wide range of magnetization and
coercivity can be obtained by the technique. The com-
posites prepared in the present work consist of a metal
phase (iron) and a semiconductor (nickel–zinc ferrite)—
2
3
ration. The precursor compounds were Ni(NO ) и 6H O,
3
2
2
Zn(NO ) и 6H O, and Fe(NO ) и 9H O, respectively.
3
2
2
3 3
2
Weighed amounts of the three nitrates were dissolved in
distilled water separately and after adding 2 drops of
concentrated nitric acid stirred for 1 h. The three solu-
tions were mixed and then stirred for 2 h. The resultant
solution was evaporated to dryness by keeping it at
200 °C for 2 h. The powder obtained after this operation
was heated at 1000 °C for 6 h. This was needed for the
purpose of calcination so that the different phases could
a)
Address all correspondence to this author.
Also affiliated with Jawaharlal Nehru Centre for Advanced Sci-
entific Research, Bangalore, 560064, India.
J. Mater. Res., Vol. 15, No. 3, Mar 2000
© 2000 Materials Research Society
683
http://journals.cambridge.org
Downloaded: 17 Feb 2015
IP address: 150.108.161.71

M. Pal et al.: Preparation of nanocomposites containing iron and nickel–zinc ferrite
be formed. The phases present in this powder were de-
contains nickel–zinc ferrite and ␣–Fe O phases, respec-
2 3
tected by taking x-ray diffractograms using a Rich Seifert
XRD 3000P (Ahrensburg, Germany) diffractometer. The
surface areas of the powders before and after heat treat-
ment were estimated to be 1.5 and 0.5 m /gm, respec-
tively. The reduction treatment of the heat-treated
powder was carried out at 650 °C for a duration of
tively. Figure 1(b) shows the x-ray diffraction (XRD)
pattern of the powder heated at 1000 °C for 6 h and sub-
sequently reduced in a stream of hydrogen at 650 °C for
10 min. It is evident that the phases present are nickel–
zinc ferrite and metallic iron, respectively. In Fig. 1(c) is
shown the x-ray diffractogram for powder heated at
1000 °C for 6 h followed by a reduction treatment in
hydrogen at 650 °C for 1 h. It is observed that only ␣–Fe
phase (metallic) is present. However, from electron mi-
croscopic investigation it will be shown later that nickel–
zinc ferrite particles are present in nanocrystalline form.
X-ray diffraction therefore does not give any detectable
signal for this phase. For further discussion on these
results we will refer to the above three powders as
specimens 1, 2, and 3, respectively. Though electron mi-
croscopic investigations of specimens 2 and 3 give un-
mistakable evidence of the presence of nano-sized
2
10 min and 1 h, respectively, in a muffle furnace wound
with Kanthal wire. A silica tube was inserted within the
furnace and hydrogen gas was flushed through the same
with a flow rate of 100 cc/min. The temperature of the fur-
nace was brought to the desired value at a rate of 100 °C/h.
After the reduction run, the gas flow was maintained and
the furnace power was shut off and it was allowed to cool
under ambient conditions. The reduction durations used
here were arrived at by a few trial experiments.
A pertinent point is the possible change of chemical
composition of the powders after each heat treatment
schedule. Mossbauer spectroscopy has emerged as a
powerful tool in recent years for identification of phases
11
in small quantities. The powders after different heat
treatments were therefore studied by Mossbauer spec-
troscopy to delineate the phases present in them. The
Mossbauer spectra of the specimens were taken in a PC-
based set-up coupled to a 1024-channel multichannel
analyzer in transmission geometry at room temperature.
57
A 12 mCi Co in Rh matrix was used as the radioactive
source. The velocity drive was calibrated with a 99.9%
pure natural iron foil of 12 ␮m thickness. Experimental
data were least square fitted to Lorentzian linewidths.
The particle size of the specimens was estimated from the
micrograph [taken in a Hitachi S-415 (Japan) scanning
electron microscope] of powders coated with a gold film.
For characterizing the ultrafine particles in samples con-
taining metallic iron, a JEM 200 CX transmission elec-
tron microscope (TEM) was used. Specimens for TEM
studies were prepared by first grinding the different heat-
treated powders under acetone by a mortar and pestle.
The powder was then dispersed in acetone and shaken in
an ultrasonic bath. One drop of acetone containing dis-
persed particles was placed on a carbon-coated copper
grid. After evaporating the liquid, the grid was mounted
onto the specimen chamber of the electron microscope.
The magnetic hysteresis properties of the different
samples were measured by a vibrating sample magn-
etometer (VSM) at room temperature. The powdered
samples were taken and mounted onto the sample cham-
ber of a VSM supplied by Princeton Applied Research
(PAR) USA (model No. 151).
III. RESULTS AND DISCUSSION
FIG. 1. XRD of different heat-treated samples with the overall com-
position 0.4 NiO, 0.4 ZnO, and 1.2 Fe O : (a) heated at 1000 °C for
2
3
In Fig. 1, the x-ray diffractograms of different samples
are shown. Figure 1(a) gives the data for the powder
heat-treated at 1000 °C for 6 h. It is seen that the sample
6
1
h, (b) heated at 1000 °C for 6 h followed by reduction at 650 °C for
0 min, and (c) heated at 1000 °C for 6 h followed by reduction at
650 °C for 1 h.
684
J. Mater. Res., Vol. 15, No. 3, Mar 2000
http://journals.cambridge.org
Downloaded: 17 Feb 2015
IP address: 150.108.161.71

M. Pal et al.: Preparation of nanocomposites containing iron and nickel–zinc ferrite
particles of nickel–zinc ferrite and ␣–Fe in them, we
TABLE I. Comparison of interplanar spacings dbk1 obtained for speci-
mens 2 and 3 with ASTM data.
have also estimated the crystallite size from the broad-
ening of the XRD peaks of the most intense lines for
Observed (nm)
ASTM
Ni0.5Zn0.5Fe O4
12
these phases by using the Scherrer equation,
.9␭
cos ␪
␣–Fe
2
Specimen 2
Specimen 3
0.2027
(nm)
(nm)
0
d =
,
(1)
0.2027
0.2027
␤
0
.2131
0.2100
0.1715
0.1417
0
0
.1704
.1348
0.1731
where d is the grain diameter, ␤ is the half-intensity
width of the relevant diffraction, ␭ is the x-ray wave-
length, and ␪ is the angle of diffraction. The particle
diameters as calculated by this method are 15.6 and
0.1433
0.1170
0.1013
0.0906
0.0827
0
.1154
0
.1064
0.1079
0.0989
0.0964
0.0758
12.8 nm for specimens 2 and 3, respectively. These are
in reasonable agreement with the values extracted
from electron micrographs as given in the following
paragraph.
Figure 2(a) is the transmission electron micrograph for
specimen 2. Figure 2(b) is the electron diffraction pattern
obtained from Fig. 2(a). The interplanar spacings, dhkl,
were calculated from the diameters of the diffraction
rings. These are listed in Table I. It is evident that the
particles shown in Fig. 2(a) consist of both nickel–zinc
ferrite and ␣–Fe. The particle sizes were measured from
the electron micrograph and the histogram was plotted.
This was fitted to a log normal distribution function as
9
delineated earlier. The mean particle diameter as esti-
mated from the histogram of particle sizes is 20 nm.
Figures 3(a) and 3(b) are the transmission electron mi-
crograph and the corresponding electron diffraction pat-
tern, respectively, obtained in the case of specimen 3.
The second column in Table I summarizes the values of
dhkl calculated from the diffraction ring diameters in
Fig. 3(b). Standard ASTM data are also listed in this
table. It is evident that in this specimen also nanometer-
sized nickel–zinc ferrite and ␣–Fe particles are present.
The median diameter of the particles in Fig. 3(a) is esti-
mated to be 15 nm.
Figure 4(a) is the Mossbauer spectrum for specimen 1.
Three sextets could be fitted to the experimental points as
shown in this figure. The sharp, outermost sextet with an
isomer shift (IS) of 0.44 mm/s, quadrupole splitting (QS)
of 0.10 mm/s, and hyperfine field of 510 kOe has been
identified as arising due to the presence of ␣–Fe O
2
3
13
phase (hematite). The other two sextets with hyperfine
fields of 447 and 417 kOe are attributed to tetrahedral
and octahedral sites, respectively, of nickel–zinc fer-
rite. The very small values of QS for the sextets indi-
cate cubic symmetry around the iron sites.
14
Figure 4(b) represents the Mossbauer spectrum for
specimen 2. Three sextets and a singlet could be fitted to
the experimental data. The sextets with hyperfine fields
of 483 and 449 kOe are attributed to the two different
sites of nickel–zinc ferrite, respectively. It is to be noted
that these internal fields are different from those obtained
for specimen 1. This indicates different Ni/Zn ratios in
these two specimens. The QS of the sextets are very
small, which shows that there is cubic symmetry around
iron atoms in both sites. The third sextet with hyperfine
field of 332 kOe is identified as arising due to the pres-
6
ence of ␣–Fe. The singlet near the zero velocity is at-
tributed to the ultrafine particles present in the specimen
giving rise to a superparamagnetic relaxation.
FIG. 2. (a) Transmission electron micrograph for specimen 2
(×100,000). (b) Electron diffraction pattern for Fig. 3(a).
J. Mater. Res., Vol. 15, No. 3, Mar 2000
Downloaded: 17 Feb 2015
685
http://journals.cambridge.org
IP address: 150.108.161.71

M. Pal et al.: Preparation of nanocomposites containing iron and nickel–zinc ferrite
FIG. 3. (a) Transmission electron micrograph for specimen 3
×100,000). (b) Electron diffraction pattern for Fig. 4(a).
(
Figure 4(c) is the Mossbauer spectrum for specimen 3.
The data could be fitted with a sextet and a doublet. The
sextet parameters with a hyperfine field of 328 kOe in-
dicates that it is due to the presence of ␣–Fe. The weak
doublet with IS of 0.38 mm/s and QS of 0.91 mm/s are
attributed to ultrafine nickel–zinc ferrite particles in the
specimen giving rise to a superparamagnetic relaxa-
FIG. 4. Mossbauer spectra of different specimens: (a) specimen 1,
b) specimen 2, (c) specimen 3.
(
particles is not contingent on the presence of free
␣–Fe O in the powder. The sub-micrometer-sized pow-
2
3
15
tion. Table II summarizes the Mossbauer parameters
obtained for different specimens.
ders (NiO, ZnO, Fe O , or Ni Zn Fe O ) are the pre-
2 3 0.5 0.5 2 4
requisites for the present method of nano-sized particle
formation. The large surface area acts as the nucleation
sites for the formation of nano-␣–Fe. It is believed that
The Mossbauer results confirm the x-ray as well as
electron diffraction data with regard to the presence of
different phases. Some comments are in order regarding
the Mossbauer and the x-ray data. Mossbauer results
clearly show that ␣–Fe O is present in specimen 1 but is
+
with the incorporation of H ions within the grains, a
tensile stress is developed, which brings about the break-
ing up of the grains. Ultimately, in specimen 3, ultrafine
particles of (Ni0.5 Zn )Fe O are produced, which give
2
3
absent in specimen 2. This is borne out by the XRD data.
The singlet observed for specimen 2 is ascribed to the
presence of nano-sized ␣–Fe particles. The negligible
value of QS indicates that the symmetry around the iron
atoms is not lost and, hence, the appearance of the singlet
in the spectra. It is to be noted that formation of ␣–Fe
0.5
2 4
rise to the doublet as seen in Fig. 4(c).
The magnetic measurements for all the specimens
show hysteresis loops. Figure 5 is a typical curve ob-
tained in the case of specimen 2. The results are sum-
marized in Table III. It is evident from this table that the
6
86
J. Mater. Res., Vol. 15, No. 3, Mar 2000
Downloaded: 17 Feb 2015
http://journals.cambridge.org
IP address: 150.108.161.71

M. Pal et al.: Preparation of nanocomposites containing iron and nickel–zinc ferrite
saturation magnetization, M , increases as the reduction
TABLE III. Summary of magnetization data for different specimens.
s
time is increased for specimen 1. It can therefore be
Saturation magnetization
(emu/g)
ascribed to the formation of ␣–Fe phase in the composite.
The coercivity value on the other hand shows a maxi-
mum in the case of specimen 2. From the XRD results it
appears that the nickel–zinc ferrite and ␣–Fe phases are
present in comparable proportions in this specimen.
Also, both these phases have nanometer-sized particles
as shown by the electron micrographs. The increase of
Specimen No.
Coercivity (Oe)
1
2
3
56.9
88.7
155.8
12.5
170
65
1 2
ր
7
T
coercivity as the particle size is reduced therefore en-
H = H
ͫ
1 − ͩ ͪ
ͬ
,
(2)
C
CO
T
hances the overall coercivity of this composite. In the
case of specimen 3, the major component present in the
composite appears to be the ␣–Fe phase as is evident
B
where HCO is the coercivity at 0° K, T is the temperature,
and T is the blocking temperature. The latter is propor-
B
from the XRD data. As a result, the value of M increases
s
tional to the volume of the ferromagnetic particle. In the
present case, the median diameter of the ferromagnetic
particle is smaller in specimen 3 than that in specimen 2.
As a result, T will be smaller in the case of specimen 3,
giving rise to a lower value of HC.
substantially. The lowering of coercivity can be ex-
plained on the basis of the theory of magnetism of
nanometer-sized ferromagnetic particles. According to
16
B
this model the coercivity, H , is given by
C
From the results presented above, it is evident that
nickel–zinc ferrite with (or without) a certain amount of
␣–Fe O , when subjected to a reduction treatment, gives
TABLE II. Summary of Mossbauer parameters for different
specimens.
2
3
rise to the formation of ferromagnetic iron particles. It is
believed that the latter form at the grain boundaries of the
ferrite particles having submicrometer-sized dimensions.
As the number of iron particles increases due to the re-
duction reaction involving hydrogen atoms, the ferrite
particle size breaks down, exposing more surface area.
This is substantiated by the fact that the particle size of
the ferrite phase decreases as the reduction treatment is
increased. (The median diameter of particles in specimen
Feature of
Mossbauer
spectrum
Specimen
No.
Heat-treatment
schedule
IS
QS
Hint
(mm/s) (mm/s) (kOe)
1
1000 ° C/6 h
3 Sextets
0.44
0.10
0.01
0.03
0.07
0.12
0.02
510
447
417
483
449
332
0
0
.29
.35
2
1000 ° C/6 h
followed by
reduction at
3 Sextets
and
0.28
0.65
0.02
0.05
0.16
6
50 ° C/10 min 1 Singlet
3
is 15 nm as compared to a value of 20 nm in the case
3
1000 ° C/6 h
followed by
reduction at
1 Sextet
and
1 Doublet
0.05
0.91
328
of specimen 2.) This is further substantiated by the
nanometer-sized nickel–zinc ferrite particles observed in
this specimen as shown in Fig. 3(a). Figure 3(b) gives the
electron diffraction pattern confirming the presence of
nickel–zinc ferrite. The particle size has been estimated
to be 20 nm. The larger surface area of the ferrite par-
ticles increases the rate of formation of iron particles at
the grain surface. By this synthesis technique it is there-
fore possible to make nanocomposites of a wide range of
magnetization and coercivity values. The nickel and zinc
apparently form an alloy system with iron.
0.39
650 ° C/1 h
In conclusion, composites consisting of nanometer-
sized nickel–zinc ferrite and ␣–Fe particles have been
prepared by a simple reduction process of micrometer-
sized nickel–zinc ferrite in the presence of ␣–Fe O . It
2
3
appears that the ferrite particles break down into nano-
meter sizes and iron particles grow at the surfaces of the
ferrite grains during this treatment.
ACKNOWLEDGMENT
This research has been supported by the Department of
Science and Technology of the Government of India.
FIG. 5. Magnetization versus magnetic field curve for specimen 2.
J. Mater. Res., Vol. 15, No. 3, Mar 2000
Downloaded: 17 Feb 2015
687
http://journals.cambridge.org
IP address: 150.108.161.71

M. Pal et al.: Preparation of nanocomposites containing iron and nickel–zinc ferrite
REFERENCES
9. B. Roy and D. Chakravorty, J. Phys.: Condens. Matter 2, 9323
1990).
(
1
2
3
4
. M.F. Crommie, C.P. Lutz, and D.M. Eigler, Science 262, 218
1993).
. B. Roy, S. Roy, and D. Chakravorty, J. Mater. Res. 9, 2677
1994).
. A.P. Alivisatos, K. Johnsson, X. Peng, T. Wilson, C. Loweth,
M. Bruchez, and P. Schultz, Nature 382, 609 (1996).
10. M. Pal, P. Brahma, and D. Chakravorty, J. Magn. Magn. Mater.
152, 370 (1996).
11. A. Okamura, S. Nakamura, M. Tanaka, and K. Siratori, J. Phys.
Soc. Japan 64(9), 3484 (1995).
12. B.E. Warren, X-ray Diffraction (Addison-Wesley, Reading, MA,
1980), p. 253.
13. N.N. Greenwood and T.C. Gibb, Mossbauer Spectroscopy (Chap-
man and Hall, London, United Kingdom, 1971), p. 241.
14. V.I. Goldanskii, V.F. Belov, N.M. Devisheva, and V.A.
Trukhtanov, Sov. Phys. JEPT 22, 1149 (1966).
(
(
. C. Mirkin, R. Letsinger, R. Mucic, and J. Storhoff, Nature 382,
6
07 (1996).
5. G. Xiao and C.L. Chien, Appl. Phys. Lett. 51, 1280 (1987).
6. S.H. Liou and C.L. Chien, Appl. Phys. Lett. 52, 512 (1988).
7. S. Roy, B. Roy, and D. Chakravorty, J. Appl. Phys. 79, 1642
15. A. Chatterjee, D. Das, S.K. Pradhan, and D. Chakravorty,
J. Magn. Magn. Mater. 127, 214 (1993).
(1996).
8
. R.H. Kodama and A.S. Edelstein, J. Appl. Phys. 85, 4316 (1999).
16. B.F. Kneller and F.F. Luborsky, J. Appl. Phys. 34, 656 (1963).
6
88
J. Mater. Res., Vol. 15, No. 3, Mar 2000
Downloaded: 17 Feb 2015
http://journals.cambridge.org
IP address: 150.108.161.71