ARTICLE IN PRESS
Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
www.elsevier.com/locate/jmmm
Magnetic properties of cobalt substituted M-type barium hexaferrite
prepared by co-precipitation
Kajal K. Mallick, Philip Shepherd, Roger J. Green
School of Engineering, University of Warwick, Coventry CV4 7AL, United Kingdom
Received 11 July 2006; received in revised form 6 September 2006
Available online 5 December 2006
Abstract
The co-precipitation and solid state methods were used in the synthesis of barium hexaferrite (BaM). Phase pure BaM was obtained
with 1, 2, 3, 5, 10, 15, 20 and 30 wt% cobalt oxide (Co3O4). The addition of Co2+/3+ ions to the BaM increased the permeability and
magnetic loss tangent to a value of 3.5 at 5% and reduced to 1 at 30% doping. With increased Co doping, Ms was reduced from 8758 emu/g, Mr increased from 11 to 40 emu/g with 3–5 wt% Co and 9 emu/g for 30% doping. Hc sharply increased from 540 to 2200 Oe
with a reduction to 280 Oe at 10 K with increasing temperature to 300 K. Tc increased from 740 to 750 K for 30% Co doping. DTA–TGA
studies of green body showed decarboxilation to occur at around 825 1C and the transformation of residual Co3O4 to Co2O3 at around
577 1C. The XRD data confirmed the Co ions substituting into Fe sites until a 10–15% doping level where the structure altered to W-type
hexaferrite. The densities of the compounds varied with doping to a maximum of 4.45 g/cm3.
r 2006 Elsevier B.V. All rights reserved.
Keywords: Electron microscopy; Magnetic property; Electrical property; Ferrite
1. Introduction
The permanent magnet and high density magnetic
recording media [1–3] are just two of the many applications
that exist for M-type hexagonal barium hexaferrite with its
stoichiometric chemical formula BaFe12O19, often denoted
as BaM. It is also known to possess high mechanical
strength as well as superior stoichiometric stability. Ferrites
and garnets are ferromagnetic oxides with dielectric and
magnetic properties that are useful for RF and microwave
applications. The parent oxide, BaFe12O19, is a hard ferrite
with a hexagonal structure belonging to the space group of
P63/mmc [4]. Due to high commercial interest in the
suitability of this compound as a material for magnetic
recording, much effort has been made to the production of
cation substituted BaM to further improve its magnetic
attributes. A variety of different cation substitutions are
possible in BaFe12O19. Divalent transition metals such as
Corresponding author. Tel.: +44 024 765 22342;
fax: +44 024 7652 4027.
E-mail address: [email protected] (K.K. Mallick).
0304-8853/$ - see front matter r 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmmm.2006.11.130
Ni2+/3+ and Co2+/3+ for Fe2+/3+ are frequently used due
to their similarity in ionic radii and electronic configurations. However, both the electrical and magnetic properties
of substituted BaM ferrites are strongly dependent on the
synthesis conditions as disproportionate charge distributions generally occur for multivalent cationic doping. There
is also a concomitant structural implication when doping
with Co2+/3+ ions influencing the magneto-dielectric
properties of this compound. Thus, cobalt doping in
particular has been the subject of many such investigations
[5–7]. This paper examines the influence of thermogravimetric, crystal structural evolution from M-type to W-type
hexaferrite, complex relative permeability and density in
hexaferrites substituted by Co2+/3+ ions. Hexaferrites are
classified into five main types depending upon chemical
formula and crystal structure; these are M, W, X, Y and Z type. The aim of introducing divalent cations such as
cobalt into M-type stoichiometry was to reduce the loss at
high frequencies whilst enhancing the permeability values.
The preparation techniques as well as the structural doping
of Co and/or Ti ions influence the magneto-dielectric
properties of this compound. For example, Mosallaies [8]
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
used barium hexaferrite mixed with cobalt oxide (Co3O4)
and barium carbonate (BaCO3), developed at Trans-Tech
Inc., and obtained a material with relative permittivity, (er0 )
and relative permeability, (mr0 ) both approximately 16, a
loss tangent of 0.001 dielectric and 0.03 magnetic at a
frequency of 500 MHz.These authors used the material in
electronic band-gap (EBG) structures and microstrip patch
antenna applications. However, they provided little experimental information as to the material preparation or
synthesis method. It is also unclear about the actual
stoichiometric composition of their material that has been
characterised. Similarly, there are many other conflicting
reports of the dielectric and magnetic properties of BaM
with little or scarce information on the ceramic or sol–gel
preparative method [9,10]. Therefore, the work presented
within this paper is hoped to show the magnetic properties
of BaM when doped with increasing quantities of cobalt.
The synthesis of samples was performed by both the
ceramic and co-precipitation methods. However, the
results of measurements are for the co-precipitation
prepared samples and the ceramic samples were done as
a comparison, results of which are not presented within the
paper. Dielectric property measurements of the compound
are still in progress and will be reported at a later date.
419
(99.999% purity, Aldrich) and Fe2O3 (99.98% purity,
Aldrich) as raw constituents, followed by further reaction
with excess Fe2O3 to obtain barium hexaferrite. Following
intermittent heating at 800 1C for 3 h and grinding (in an
agate mortar and pestle) various samples were formed and
different percentages of Co3O4 by weight were added to the
phase pure BaFe12O19. Following further intermittent
heating at 800 1C for 3 h and grinding the final material
samples were pelletised using a Specac uniaxal press to
6000 psi producing disk samples of 16 mm in diameter.
Sintering of these pellets was carried out at temperatures
ranging from 900 to 1200 1C, the latter being the final
sintering temperature. The barium cobalt ferrite composition produced by both methods was subjected to identical
temperature regimes and time intervals.
2.2. Density measurements
The density of the samples from the green to a sintered
state was measured using the well known liquid displacement Archimedes’ technique. The measurements were
carried out to record the change in density following
various sintering treatment protocol.
2.3. X-ray diffraction measurements
2. Experimental
2.1. Synthesis
A chemically reliable co-precipitation route as well as the
classic solid state synthesis method from relevant oxides
was used in the preparation of a stoichiometric composition of BaM.
The solution route involved a slow and drop wise
addition of excess ammonium hydroxide (pH ¼ 12) to the
aqueous solution consisting stoichiometric amounts of
nitrates of Ba and Fe (99.99% purity, Aldrich) that
resulted in the co-precipitation of a gel like cake. The cake
was thoroughly washed with deionised water and purified
ethanol, dried at 105 1C and was subsequently heated to
various temperatures.
The ceramic method involved firstly the formation
of stoichiometric monoferrite, BaFe2O4 using BaCO3
Chemical composition analysis was carried out by
powder X-ray diffraction (XRD). XRD patterns for
samples of 0%, 3%, 15% and 30% Co3O4 doping by
weight treated at various temperatures and times were
recorded in the region of 2y ¼ 10–801 with a step scan of
0.021/min on a Philips diffractometer (Model PW1710)
using CuKa radiation. Automated powder diffraction
software package that included both standard International Centre for Diffraction Data (ICDD) and calculated
ICSD diffraction files was used to match the evolving
phases. Cell parameters were calculated and further refined
using linear regression procedures (Philips APD 1700
software) applied to the measured peak positions of all
major reflections up to 2y ¼ 901.
The well-known Scherrer formula [11] was used to
determine the crystallite size from the line broadening of
diffraction profile of the strongest peak. The formula,
Fig. 1. (a) Parallel plate test fixture measurement method. (b) Equivalent circuit model.
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
420
Fig. 2. Thermogravimetry (TGA) and differential thermal analysis (DTA) of (a) as synthesized co-precipitated BaM cake and (b) BaM cake with 3% Co
doping, (c) BaM cake with 10% Co doping and (d) BaM cake with 30% Co doping by weight.
where D ¼ average size of the crystallites, k ¼ Scherrer
constant, (k ¼ 0.9). l ¼ wavelength of radiation (lCuKa ¼
1.54056 Å), h1/2 ¼ peak width at half height (full width half
maxima, FWHM measured in radians) and y corresponds
to the peak position measured in radians. The measured
values were corrected for the effects caused by instrumental
broadening.
2.4. Scanning electron microscopy
Determination of the morphology was performed using
scanning electron microscopy (SEM). A Philips Cambridge
Stereoscan was used to determine the phase pure and Co
ion doped barium hexaferrite particles as well as to observe
related microstructural features of the ferrite particulates.
Fig. 3. Density, temperature, and time variations.
2.5. Complex relative permeability measurements
excluding the effects of the machine broadening to
minimise errors, is given below:
D¼
kl
cos y,
h1=2
(1)
The high frequency magnetic properties were measured
with an Agilent Technologies Limited, Impedance/Material Analyser E4991A, with a 16454A test fixture for
magnetic material measurements. The frequency range
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
Fig. 4. X-ray diffraction patterns of sample co-precipitation method
compound sintered at 1200 1C/2 h, Standard JCPDS File (27-1029) and (a)
the pure BaM (b) BaM compound with 1% Co doping, (c) BaM
compound with 2% Co doping, (d) BaM compound with 3% Co doping.
was 1 MHz–1 GHz and measured at a temperature of
25 1C.
The test fixture uses the inductive method where the
device under test (DUT) is shaped into a toroidal core and
is wrapped with wire, the complex relative permeability
calculated from the inductance values at the end of the
toroid. The ideal impedance of the test fixture, when no
DUT is present, is given by Zss, and the residual impedance
Zres can be computed from the measured impedance Zsm.
Z res ¼ Zsm Z ss .
(2)
Errors due to residual impedance can be compensated by
the calibration procedure, using the SHORT state, with no
DUT present in the test fixture. The impedance after error
correction, Zcomp, can be computed from the measurement
impedance Zm, with a DUT mounted in the test fixture, as
in Fig. 1. The corrected impedance was calculated as
follows:
Z comp ¼ Z m Z res .
(3)
421
Fig. 5. X-ray diffraction patterns of sample co-precipitation method
compound sintered at 1200 1C/2 h, Standard JCPDS File (27-1029): (a) the
pure BaM; (b) BaM compound with 3% Co doping; (c) BaM compound
with 5% Co doping; (d) BaM compound with 10% Co doping.
Assuming that Zss consists only of inductance elements i.e.
e
m Zss ¼ jo o ho ln
.
(4)
a
2p
The complex relative permeability of the DUT can be
computed thus,
2pðZ m Z sm Þ
þ 1,
mr ¼
(5)
jomo h ln bc
where
m ¼ mo ðm0r jm00r Þ.
(6)
A thorough examination of the measurement procedure
illustrated above can be found in [12].
2.6. Vibrating sample magnetometer measurements
A vibrating sample magnetometer (VSM) from Oxford
Research Instruments, for magnetic measurement was used
and data was analysed by proprietary Oxford Object Bench
software. The measurement of the magnetic properties of
materials was performed as a function of magnetic field,
temperature, and time. Pellets of a 3 mm diameter were
ARTICLE IN PRESS
422
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
Fig. 6. X-ray diffraction patterns of sample co-precipitation method
compound sintered at 1200 1C/1 h, Standard JCPDS File (27-1029) and
(19-0098): (a) the pure BaM; (b) BaM compound with 10% Co doping; (c)
BaM compound with 15% Co doping; (d) BaM compound with 20% Co
doping; and (e) BaM compound with 30% Co doping.
placed inside the VSM for samples containing 0%, 3%,
5%,10%, 20% and 30% Co doping. The resulting hysteresis
loops provided the relationship between magnetisation, M,
and the applied field, H, at temperatures of 263, 73 and
27 1C by means of a helium cooled 12 T super-conducting
magnet. The parameters extracted from the hysteresis loop
that are most often used are the saturation magnetisation
(Ms), the remanence (Mr), the coercivity (Hc), the squareness
ratio (SQR), which in turn is related to the slope at Hc, and
the ordinary relative permeability (mrord). In addition, the
VSM was used to determine the Curie point of the material
for 0%, 5%, 10% and 30% Co doped BaM samples.
3. Results and discussion
3.1. Thermogravimetric (TGA) and differential thermal
analysis (DTA)
Fig. 2(a–d) shows the DTA–TG plots, for the green
unsintered compound containing a mixture of BaCO3,
Fig. 7. Structural representation of BaFe12O19 (BaM) compound along
the (1 0 0) axis (small filled, medium and large circles ¼ O, Fe and Ba,
respectively).
Fe2O3 and Co3O4 doping of 0%, 3%, 10% and 30% by
weight. For the BaM sample, the weight loss is incremental
and an overall loss of about 4% is indicated. There is an
appearance of an endotherm in the DTA curve at 807 1C.
The peak is attributed to the decarboxilation of BaCO3,
reported to take place at 1055 1C for pure carbonate [13]
and around 800 1C for the mixture of carbonate and iron
oxide [14]. The completion of the formation of the
hexaferrites is indicated at around 1050 and 1011 1C for
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
423
3.3. Crystal structure and particle morphology
Fig. 8. 100% peak shifting with Co doping.
Table 1
Summary of powder X-ray diffraction measurements and average
crystallite size for samples sintered at 1200 1C
Doping Co (% by
weight)
100% Peak
position (deg)
FWHM
(deg)
Average
crystallite
size(D) (mm)
JCPDS:27-1029
0
1
2
3
5
20
30
32.1730
32.1785
32.2129
32.1810
32.2922
32.1822
35.4972
35.4950
—
0.1998
0.2169
0.2084
0.2129
0.3010
0.2048
0.1978
—
4.69
4.32
4.50
4.41
3.11
4.76
4.93
JCPDS:19-0098
34.6050
—
—
the W-type ferrites. The exotherm in the DTA curve at
577 1C is associated with residual Co3O4 to Co2O3
transformations.
3.2. Variation of density
Fig. 3 shows the isothermal and isochronal variation of
density for doped BaM at various doping regimes over
temperatures and time intervals shown. For the undoped
BaM, after the final sintering of 1200 1C for 2 h the density
was found to be 4.22 g/cm3, density then peaked at 4.45 g/
cm3 with 1% doping. With increasing level of doping there
was a general decrease in density values, whereupon at
20%, the value was found to be 3.5 g/cm3, and then
increasing sharply to 4.21 g/cm3 for 30% doping. During
sintering regimes the density varied little with time,
temperature and doping levels, all remaining at around
4.2 g/cm3 level.
Figs. 4–6 show the typical XRD patterns of the sintered
BaM (0% doping of Co3O4) together with Co3O4 doped
BaM at 1%, 2%, 3%, 5%, 10%, 15%, 20% and 30% by
weight. The figures also include the standard (JCPDS no:
27-1029) [15] for barium hexaferrite (BaFe12O19) and
(JCPDS no: 19-0098) [16] for barium cobalt iron oxide
(BaCo2Fe16O27). It can be seen from the Fig. 4(a) that the
initial BaM compound is of a pure single phase and agrees
well with the standard 27-1029, including similar relative
intensity profiles for the first three strong peaks. The peaks
were indexed as a primitive hexagonal cell with space
group, P63/mmc (194) with the refined lattice parameter
values of a ¼ b ¼ 5.895 Å and c ¼ 23.199 Å. These values
agree well with the published literature data for BaM
[17–19]. A schematic structural representation along the
(1 0 0) axis projection of the parent phase, BaFe12O19 is
shown in Fig. 7. The compound crystallizes in the
hexagonal space group, P63/mmc consisting 64 ions per
unit cell on 11 different symmetry sites. There are 24 Fe3+
atoms distributed over 5 symmetry sites with 3 octahedral,
1 tetrahedral and 1 bipyramidal site as described by
Collomb et al. [20] whereupon they perform a neutron
diffraction study of the crystal structure and cobalt
locations within a W-type hexagonal ferrite. The site
occupancy during substitution is largely dependent on the
relative ionic radii of the substituent atoms. From the data
published by Shannon [21], only Co3+ (rCo ¼ 0.545 Å) is
expected to replace Fe3+ (rFe ¼ 0.645 Å) due to their
similarity in ionic size.
Both M and W-type ferrites have similar hexagonal
structure and as a result most peaks in the XRD plots are
very similar. In fact it is necessarily difficult to distinguish
between these two structures. Fig. 4(a) corresponding to
the ‘‘initial composition’’ is phase pure, single phase and
matches, exactly the peaks corresponding to the standard
JCPDS file (27-1029) for BaFe12O19. The relative intensity
values for the first three strongest peaks, namely (107),
(114) and (008) are also similar to the standard pattern as
well as the literature data published by Mandal et al [22].
From 5% to 10%, where, if one allows for the background
noise the structure resembles W-type particularly with the
increased intensity of the main W-type peaks in the region
of 25–401 as well as between 50 and 651. The peaks are very
similar to those reported by Mandal et al. In addition to
the XRD results the magnetisation over temperature
results for the initial composition of BaFe12O19 shows a
very smooth curve with increasing temperature until the
Curie point at 725 K. Had the composition been multiphase then a knee would be present in the curve as seen in
the 5% and 10% doped samples where there is a
transitional phase from M-type to W-type. The 100%
peak in 27-1029 (107) is present in 0–3% with relative
intensity also matching closely to the standard pattern.
However, as the doping is increased there is a clear
indication of the 100% peak intensity belonging to the
ARTICLE IN PRESS
424
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
Fig. 9. Representative micrograph of: (a) Pure BaM; (b) BaM compound with 2% Co doping; (c) BaM compound with 15% Co doping; and (d) BaM
compound with 30% Co doping.
W-type (19-0098) BaFe16Co2O27 (1 1 6) becoming more
intense thus suggesting strongly the appearance of the Wtype phase. Moreover, in view of this, this would suggest
the transformation occurs from 5% to 10% doping. With
30% doping, the structure is somewhat different from the
W-type, although there is still a strong possibility of a shift
in all the peaks (due possibly to preferred orientation). We
are unable at this time to identify whether it is a new phase
or a structural variation of the same W-type, within
perhaps a slightly different space group. Within the 5%
doped XRD plot there appears doublets that are a clear
indication of a changeover in structure.
From the FWHM study of the 100% peak (1 0 7) in the
0%, 1%, 2%, 3%, and 5% doping (27-1029) and the 100%
peak (116) in the 20 to 30% doping (19-0098), which can be
seen in Fig. 8. The 100% peak shifting, FWHM values and
crystallite size, D are summarised in Table 1.
This phase conversion is much more apparent as the
level of doping is increased further. Here, this new phase
now replaces the old parent phase and is identified as a
compound based on a stoichiometric composition corresponding to JPDCS no; 19-0098. Clearly, there is a limit as
to how much Co ions can be incorporated in the parent
phase before the parent phase dissociates, reacts with the
excess Co3O4 and, perhaps simultaneously converts to this
new phase.
Fig. 9(a–d) show SEM representative micrographs of the
phase pure BaM (0%) and 2%, 15% and 30% doping of
Co3O4 pellets after sintering at 1200 1C for 2 h. The
particles were shown to have the following dimensions,
0% doping between 5 and 17 mm, 2%, 15% and 30%
doping between 3 and 5 mm.
3.4. High frequency complex relative permeability
measurements
The real component of the complex relative permeability
(m0r ) was found to be frequency invariant over the range of
50 to 1000 MHz. The plot of mean average value of the real
component of the complex relative permeability versus
percentage Co doping is shown in Fig. 10(a). Within the
sintering regimes of 900–1050 1C there was only a very
slight increase in the relative permeability regardless of Co
doping level, from 1.252 to 1.36. However, on a sintering at
1200 1C for 2 h the m0 r continued to remain constant over
frequency. The pure BaM sample showed a m0 r value of
1.31. This increased linearly to a peak of 3.48 on a doping
level of 5%, then falling to 1.76 with 20% and subsequently
to 1.32 at 30%. This trend corresponds to the phase change
determined by the XRD data indicating the M-type BaM
being no longer able to accommodate any surplus Co ions
between 5% and 10% doping. It is here that the W-type
structure is thought to form, indicated by the XRD plots.
From the percentage of Co doping it may be possible to say
that the maximum value of m0 r would occur between 5%
and 10% doping.
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
Fig. 10. (a) Mean average value of real part of complex relative
permeability over time, temperature, and frequency and Co doping; (b)
Typical plot of real part of complex relative permeability over time,
temperature, and frequency of samples sintered at 1200 1C for 2 h.
Fig. 10(b), shows a typical plot of samples sintered at
1200 1C for 2 h, displaying a slightly increasing response of
m0 r over frequency in the range of 50–1000 MHz.
The mean average value of the imaginary component
of the complex relative permeability (m00r ) is plotted in
Fig. 11(a) against percentage Co doping by weight,
sintering time and temperature. At the lowest sintering
temperature of 900 1C for 4 h, the m00r was relatively constant
regardless of doping levels at 100 103. Further sintering
at 1000 and 1050 1C reduced the m00r to approximately
50 103 and this value continued to remain stable
regardless of doping levels over frequency. Upon a final
sintering regime of 1200 1C for 2 h the initial m00r value at 0%
doping was 160 103. At 5% doping, the value peaked
sharply to 504 103. With increasing doping (5–20%),
the value fell to 60 103 and to 15 103 at 30%.
425
Fig. 11. (a) Mean average value of imaginary part of complex relative
permeability over time, temperature, and frequency and Co doping. (b)
Typical plot of imaginary part of complex relative permeability over time,
temperature, and frequency of samples sintered at 1200 1C for 2 h.
A typical plot of samples for all doping levels sintered at
1200 1C is shown in Fig. 11(b), displaying a spike in m00r at
50 MHz then falling to approximately 200 103 at
400 MHz and gradually rising thereafter until 1000 MHz.
The magnetic loss tangent, tan dm is the ratio of the
imaginary part to the real part of the complex relative
permeability. The tangent of the phase angle d of the
complex relative permeability is defined in terms of m0r
and m00r .
tan dm ¼
m00r
.
m0r
(7)
Thus, Fig. 12(a) that shows a plot of the mean average
value of m00r over frequency against percentage Co doping
by weight, sintering time and temperature and Fig. 12(b)
showing a typical plot of samples sintered at 1200 1C are
the quotient of m00r and m0r (loss tangent tan dm as referred to
by Eq. (7)) for isothermally and isochronally treated
ARTICLE IN PRESS
426
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
Fig. 12. (a) Mean average value of the magnetic loss tangent over time,
temperature, and frequency and Co doping. (b) Typical plot of magnetic
loss tangent over time, temperature, and frequency of samples sintered at
1200 1C for 2 h.
samples, and as such this loss parameter follows the same
trends as the m00r as the m0r values are elevated at 50 MHz,
and reduce to a minimum at 400 MHz and begin to rise
once more until the upper frequency range of the tests
performed is 1000 MHz. Unfortunately, as the m0r is
enhanced the loss tangent increases, therefore, for certain
applications such as resonator or microstrip antennas this
compound is not suitable. However, for other applications
such as absorbers the BaM doped compound becomes a
suitable compound for use in this area.
3.5. Magnetisation measurements
The hysteresis curves in Fig. 13(a–f) for samples with
0%, 3%, 5%, 10%, 20% and 30% Co doping, sintered at
1200 1C for 2 h, show the magnetic data measured at 10,
100, 200 and 300 K are summarised in Table 2. Measure-
ments from the VSM show that the intrinsic saturation Ms,
is achieved when a high H field is applied. In each sample,
the lower the temperature the higher is the value of
saturation Ms. Additionally, with the increase in doping,
regardless of temperature, the saturation magnetisation is
reduced. For example, at 300 K for pure undoped BaM the
saturation magnetisation had a value of 60.1 emu/g. Upon
doping with 30% Co that value reduced to 55.1 emu/g.
The remanence magnitude can be extracted from the
hysteresis loop at the intersections of the loop with the
vertical magnetisation axis. The pure BaM samples
recorded an increase in remnant magnetisation Mr with
the increase in temperature, from 11.5 emu/g at 10 K to
14.6 emu/g at 300 K. With increased doping, the value at
10 K is enhanced greatly from 11.5 emu/g (at 0% doping)
to 39.5 emu/g at 30% doping. However, an increase in
temperature to 300 K the remnant magnetisation resulted
in a large reduction of value below that for the pure BaM.
SQR for these samples is denoted by the ratio of (Mr/
Ms). This is essentially a measure of squareness of the
hysteresis loop. Compared to this relatively low value, in
general however, large SQR values (in the region of 0.8) are
preferred in many applications such as magnetic recording
media of high density. The pure BaM samples recorded an
increase in SQR with the increase in temperature, from
0.131 at 10 K to 0.242 at 300 K. With increased doping of
Co, at 10 K the value is enhanced greatly from 0.131 (at 0%
doping) to 0.673 at 30%. However, with a temperature
increase to 300 K of the doped samples the SQR is greatly
reduced to a value below the one for the pure BaM.
The data for all samples is summarised in Table 2. For
the pure BaM samples, the coercivity (denoted here as Hc)
results obtained agree well with the published literature
[23,24]. The recorded results showed an increase in
coercivity with the increase in temperature, from
540.06 Oe at 10 K to 860.57 Oe at 300 K. With a further
increase in doping at 10 K the value was enhanced greatly
from 540.06 Oe (at 0% doping) to 2213.78 Oe at 30%.
However, as seen in the previous results, an increase in
temperature to 300 K resulted in a reduced coercivity value
to that observed for pure BaM. Interestingly, viewed in
Fig. 13 and summarised in Table 2, it can be seen that the
SQR and coercivity increase as the measurement temperature is increased on the pure undoped BaM samples. The
rate of increase is somewhat lessened for the 3% doping
samples and a reversal to decreasing SQR and coercivity is
observed for 5% and 10%, where the W-type structure is
apparent. Upon 20% and 30% doping the rate of decrease
in SQR and coercivity becomes larger still. The temperature dependency of BaM when doped with Co ions is
changed and finally reversed on a dopant level between
3%wt and 5%wt. The trend between rate of change in SQR
and coercivity is similar to the value of density in the
samples. It cannot be determined whether there is, in fact a
connection between temperature dependency of the SQR
and coercivity and the density of the samples that were
tested.
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
427
Fig. 13. (a) Hysteresis curve for pure BaM. (b) Hysteresis curve for BaM doped with 3% Co. (c) Hysteresis curve for BaM doped with 5% Co. (d)
Hysteresis curve for BaM doped with 10% Co. (e) Hysteresis curve for BaM doped with 20% Co. (f) Hysteresis curve for BaM doped with 30% Co.
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
428
Table 2
Summary of magnetisation measurements for samples sintered at 1200 1C
Temperature (K)
Saturation (Ms) (emu/g)
Percentage Co doping (% by weight)
0
10
100
200
300
87.688
81.675
71.500
60.175
3
5
10
20
30
77.921
74.475
69.567
60.103
86.547
86.271
79.456
64.043
72.960
72.330
66.172
58.992
57.550
52.230
51.687
50.612
58.750
58.931
58.339
55.171
10
100
200
300
Remnant (Mr) (emu/g)
11.560
21.349
12.343
17.550
12.647
17.535
14.634
17.556
22.283
14.822
10.455
8.409
20.925
13.096
9.171
6.228
28.306
26.095
16.788
12.617
39.566
38.068
17.810
9.493
10
100
200
300
SQR (Mr/Ms)
0.131831
0.151087
0.176783
0.242626
10
100
200
300
Coercivity (Hc) (Oe)
540.06
785.60
604.67
615.14
685.30
612.68
860.57
1079.4
0.273867
0.235650
0.251987
0.291999
0.257467
0.171807
0.131582
0.131302
573.94
423.40
398.67
378.27
0.286801
0.181059
0.138593
0.105574
531.96
272.65
174.94
102.34
0.491859
0.499617
0.324858
0.249289
2396.1
1686.1
786.56
476.19
0.673464
0.645987
0.305285
0.172065
2213.78
1310.48
519.651
218.330
at 750 K. The data for both the 5% and 10% doping would
appear to suggest presence of a multiphase material as
already indicated by the XRD results. In contrast, the
compound with a 30% doping, a single phase material is
present with a Tc of 745 K. These highly doped samples
exhibited an overall higher magnetisation level and a
higher Curie point.
It is important to stress that, in the present investigation;
the magnetic moments are thought to be partially aligned
within the magnetic domains in the samples. This is due to
the fact that the samples were not sintered in the presence
of a magnetic field. Hence, the expression of the Tc values
in this case can be explained in terms of the number of
Fe3+–O2–Fe3+ effective interactions. This is in agreement
with a similar explanation provided by Gilleo [25].
Fig. 14. Curie point measurement for BaM samples with 0%, 5%, 10%
and 30% Co doping by weight.
3.7. Ordinary relative permeability
3.6. Curie point measurement
The ordinary relative permeability can be described as
follows:
The variations of magnetisation with temperature for
samples doped with 0%, 5%, 10% and 30% by weight Co
and sintered at 1200 1C for 2 h are shown in Fig. 14.
The Curie point or temperature (Tc), measured in the
VSM, was found to be 725 K for undoped samples. The
results are in agreement with Tc measurements reported by
the ASM International [23] and Ram [24]. For the samples
with 5% doping, there appears to be a knee at 692 K;
otherwise Tc was found to be at 745 K. With 10% Co
doping, two knees are present: one at 690 K and the other
mord ¼
~max
B
~ max
mo H
or
7:96 105
~max
B
,
~ max
H
(8)
where Bmax, magnetic flux density is taken at Ms (Tesla)
and Hmax, is the applied field (A/m). The results relating to
ordinary permeability at different temperatures, obtained
from the hysteresis plots using Eq. (8), are shown in Fig.
13(a–f) and are summarised in Table 3. In general, the
trend of the ordinary relative permeability for the doped
samples is similar to our earlier discussion for the
ARTICLE IN PRESS
K.K. Mallick et al. / Journal of Magnetism and Magnetic Materials 312 (2007) 418–429
Table 3
Summary of ordinary relative permeability measurements for samples
sintered at 1200 1C
Temperature (K)
influence on altering the structural makeup, and therefore
the magnetic and electrical characteristics of the final
compound and hence the device.
Ordinary relative permeability (mrord)
Acknowledgements
Percentage Co doping (by weight)
10
100
200
300
429
0
3
5
10
20
30
92.58
86.23
75.49
63.53
80.31
76.76
71.70
61.95
86.59
86.31
79.50
64.07
66.76
66.18
60.55
53.98
50.63
45.95
45.47
44.53
62.47
62.66
62.03
58.66
saturation magnetisation, remnant magnetisation, coercivity and squareness. Here, the ordinary relative permeability
is at a maximum of 92.58 at lower temperatures and at
lower Co doping levels. These values are significant for
applications involving an ac current such as cores of coils,
ferrite rods, RF chokes and similar such applications.
4. Conclusions
The variation of magnetic properties of Co2+/3+ doped
barium hexaferrite BaFe12O19 with frequency have been
presented. These include relative permeability measurements over the frequency range of 1 MHz–1GHz at room
temperature. The addition of Co2+/3+ ions to the Barium
hexaferrite increased the permeability from a value of
1.2–3.5 at 5% doping which subsequently was reduced to 1
at 30%. The value of relative permeability appears to be
related to the evolution of characteristic crystallographic
phase in this compound, where M phase is converted to Wtype at higher dopant level (30 wt%). XRD study of the
BaM doped with 5% and 10% Co doping confirmed the
presence of multiple phases. This is corroborated by the
occurrence of the knees within the M versus T plots for the
Curie point (Fig. 14). The loss tangent is similarly affected.
With increased Co doping, the saturation magnetisation
Ms was reduced from 87 to 58 emu/g. The remnant
magnetisation Mr increased from 11 to 40 emu/g with
3–5 wt% Co and 9 emu/g for 30% doping. The coercivity
Hc sharply increased from 540 to 2200 Oe for 0% and 30%
doping, respectively at 10 K, with a reduction of the 30%
doping to 280 Oe with increasing temperature to 300 K.
Very close temperature dependency on the magnetic
properties was observed, the SQR and coercivity appear
to be corresponding to the density response over percentage Co doping. The Curie point Tc increased from 740 to
750 K for 0–30% Co doping, respectively.
The phase evolution of the compound from an M-type
BaM to a W-type barium cobalt ferrite oxide is significant
when a higher level of doping is considered. The magnetic
properties are affected greatly throughout the sintering and
preparation process. When designing high density recording media, radio frequency absorbers or surface mount
chip devices, it is important to take into consideration the
level of multivalent Co in Co3O4 as it has a profound
The authors would like to thank Mr. Martin Davies for
his assistance in the Materials Preparation and Microscopy
Laboratory, School of Engineering, University of Warwick, and for Electron Microscopy and XRD of the
dielectric materials. Our thanks are due to Dr. Martin Lees
for VSM measurements and Mr. David Hammond for the
use of the TG-DTA equipment at the Department of
Physics, University of Warwick.
References
[1] H. Pfeiffer, R.W. Chantrell, P. Görnert, W. Schüppel, E. Sinn,
M. Rösler, J. Magn. Magn. Mater. 125 (1993) 373.
[2] J. Smit, H. P. J. Wijn, Ferrites, Philips Technical Library, Eindhoven
1961.
[3] W. Buchner, R. Schliebs, G. Winter, K. H. Buchel, Industrial
Inorganic Chemistry, V.C.H. Weinheim, 1989.
[4] J. Smit, H. P. J Wijn, Ferrites: physical properties of ferromagnetic
oxides in relation to their technical applications, Philips Technical
Library, 1959.
[5] R. Carey, J. Appl. Phys. 75 (1991) 6789.
[6] X. Batlle, J. Appl. Phys. 70 (1991) 1614.
[7] X. Batlle, J. Appl. Phys. 74 (1993) 3333.
[8] Hossein. Mosallaei, K. Sarabandi, IEEE Trans. Antennas Propagation AAP52 (2004) 1558.
[9] Y.S. Homg, C.M. Ho, H. Y Hsu, C.T. Liu, J. Magn. Magn. Mater.
279 (2004) 401.
[10] G. Benito, M.P. Morales, J. Requena, V. Raposo, M. Vazquez, J.S.
Moya, J. Magn. Magn. Mater. 234 (2001) 65–72.
[11] H.P. Klug, L.E. Alexander, X-Ray Diffraction Procedures for
Polycrystalline and Amorphous Materials, Wiley, New York, 1967.
[12] Begley Shelley, Application note 1217-1, Basics of Measuring the
Dielectric Properties of Materials, Agilent Technologies, September
2001.
[13] C. Duval, Inorganic Thermogravimetric Analysis, Elsevier, London,
1963, pp. 533.
[14] G.C. Bye, C.R. Howard, J. Appl. Chem. Biotechnol. 21 (1971) 319.
[15] JCPDS File Number. 27-1029, International Centre for Diffraction
Data (ICDD).
[16] JCPDS File Number. 19-0098, International Centre for Diffraction
Data (ICDD).
[17] R. Routil, Barham. J. Chem. 52 (1974) 3235.
[18] W.D. Townes, J.H. Fang, A.J. Perrotta, Zeitschrift fuer Kristallographie, Kristallgeometrie, Kristallphysik, Kristallchemie 125 (1967)
11.
[19] E.F. Bertaut, A. Deschamps, R. Pauthenet, Comptes Rendus
Hebdomadaires des Seances de l’Academie des Sciences 246 (1958)
2594.
[20] A. Collomb, B. Lambert-Andron, J.X. Boucherle, D. Samaras, Phys.
Stat. sol. (A) 96 (1986) 385.
[21] R.D. Shannon, Acta Crystallogr. A 32 (1976) 751.
[22] S.K. Mandal, K. Singh, D. Bahadur, J. Mater. Sci. 29 (1994) 3738.
[23] ASM Handbook–‘‘Properties and Selection: Nonferrous Alloys and
Special-Purpose Materials.’’ vol. 2, ASM International, Materials
Park, OH, 1990.
[24] S. Ram, H. Krishnan, K.N. Rai, K.A. Narayan, Japanese J. Appl.
Phys. 28 (4) (1989) 604.
[25] M.A. Gilleo, Phys. Rev. 109 (1958) 777.