Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

ARTICLE IN PRESS

Journal of Magnetism and Magnetic Materials 317 (2007) 61–67


www.elsevier.com/locate/jmmm

Different annealing treatments for improvement of magnetic and


electrical properties of soft magnetic composites
H. Shokrollahi, K. Janghorban
Materials Science and Engineering Department, School of Engineering, Shiraz University, Shiraz, Iran
Received 27 December 2006; received in revised form 27 March 2007
Available online 20 April 2007

Abstract

This paper investigates the effect of two steps, annealing and magnetic annealing, on the magnetic and electrical properties of iron
powder particles with high purity used in soft magnetic composite materials. It was found that at low frequencies (o10 kHz) the
magnetic loss in the high-temperature magnetic annealed state is smaller than that in the low-temperature magnetic, magnetic annealed
and without annealing state. On the contrary, the magnetic loss in the high-temperature magnetic annealed state is larger at high
frequencies (410 kHz) than the others. The results showed that magnetic loss of annealed powder was smaller than that of unannealed
powder after 100 h milling process. Also, the magnetic loss of samples prepared by two steps milling and annealing process was smaller
than samples with one-step process. The magnetic and electrical properties of the cylindrical samples were measured by an LCR meter.
Microstructures of the samples were characterized by SEM and XRD.
r 2007 Published by Elsevier B.V.

Keywords: Soft magnetic composite; Magnetic annealing; Two steps annealing; Magnetic loss

1. Introduction Annealing is required to minimize the deleterious effects of


cold work on the magnetic performance of the core
To achieve low power loss in a soft magnetic material for material. Generally, annealing can be divided into three
AC applications, the soft magnetic cores must meet the categories: thermal annealing, magnetic field annealing and
following three requirements: (1) lower coercive force, (2) thermal–magnetic field annealing.
greater refinement of eddy current producing region, and The mechanism operating to achieve improvement of
(3) higher electrical resistance. Insulated iron powder magnetic properties is closely related to better crystallinity
materials, sometimes also called soft magnetic composites and reduction of grain boundaries caused by magnetic
(SMCs), have isotropic magnetic behavior, improved high- annealing [13]. Annealing a sample will change the domain
frequency performance and can be compacted to 3D and structure due to stress release and hence would have an
complex-shapes using the established PM compaction effect on the anomalous loss.
route [1–6]. During compaction internal stresses are generated in the
To produce a powdered compressed magnetic core material. The hysteresis loss is partly due to stresses
(SMC), high pressures should be applied. Because of the introduced in the material at compaction which can impede
presence of polymer materials, sintering process cannot be domain wall movement. In practice, it is energetically
used, while residual stresses which have been induced favorable for domain walls to pass through certain
during milling process can deteriorate magnetic properties. imperfections such as stressed regions. Therefore in order
For this reason, annealing process for the elimination of to reduce hysteresis in the iron-based composite, a stress
residual stresses in the compaction step is essential [7–12]. relieving low-temperature heat treatment most often
follows the compaction [2]. The residual stresses tend to
Corresponding author. minimize by thermal motion of the atoms. Heat treatment
E-mail address: shokrola@shirazu.ac.ir (H. Shokrollahi). provides high uniformity film, low-volume fraction of

0304-8853/$ - see front matter r 2007 Published by Elsevier B.V.


doi:10.1016/j.jmmm.2007.04.011
ARTICLE IN PRESS
62 H. Shokrollahi, K. Janghorban / Journal of Magnetism and Magnetic Materials 317 (2007) 61–67

defects and reduces the distortion within particles and so homogenized mixture. After the evaporation of the solvent,
the coercive force reduces and the magnetic permeability the coated powders were obtained. Following drying, the
increases. powders were uniaxially cold compacted at 800 MPa into a
The crystalline quality is significantly improved by cylindrical die with a diameter of 12 mm (die wall lubricant
application of the external field. Magnetic field annealing was glycerin). The samples were cured in air at 200 1C for
is much more effective in enhancing the crystallinity of 60 min.
samples compared to non-field heat treatment. Magnetic
field is used to homogenize the magnetic properties. The 2.2. Magnetic annealing
enhanced properties may be explained by improved
crystallinity and reduced density of grain boundaries, In this experiment, for reduction in shape anisotropy, the
which is caused by magnetic field annealing [13]. In the iron powders with high purity were milled for 50 h and for
present magnetic field, domain boundaries pass some reducing the undesirable effects of residual stresses, the
pinning sites and magnetic domains will be larger and powders were annealed at 550 1C for 1 h at argon atmo-
domain growth will be easier. Also, with respect to sphere (average particle size o50 mm). Coated powders
anisotropic properties, effective permeability in the mag- were prepared as explained before. For investigation of the
netic field direction increases. effect of magnetic annealing on the magnetic properties,
This paper focuses on the effect of different annealing three different treatments were performed. A magnetic flux
treatments including two steps annealing, and magnetic density (0.5 T) was applied to the sample at room
field annealing on the magnetic properties of soft magnetic temperature as magnetic annealing and a similar magnetic
composite materials. field was applied to the samples at 180720 and 280720 1C
as low- and high-temperature magnetic annealing for 300 s.
2. Experimental procedures In these experiments the sample holder, containing the
sample, was positioned at the center of the magnetic field,
The iron powder was supplied by Merck with particle with the sample parallel to the applied field. For better
size o150 mm. The purity of Fe was above 98% containing comparison a sample was produced without any annealing
0.02% C, 0.01% Cu, 0.002% Pb, 0.005% Zn and some treatment. The magnetic loss, effective permeability and
oxides. Two different experimental methods used in this electrical properties of the cylindrical samples were
work are explained as follows. measured by an LCR meter (HIOKI 3532 LCR).
Morphologies of the powders and samples were character-
ized using scanning electron microscopy (Cambridge
2.1. Two steps annealing
Stereoscan). X-ray diffraction experiments were done by
Brucker Advance 8 diffractometer using CuKa-radiation.
Iron powders were milled under different conditions.
Series 1 iron powder were milled for 100 h. Series 2 iron
3. Results and discussion
powders were milled for 50 h and then annealed under
argon atmosphere at 550 1C for 1 h, and milled again for
3.1. Two steps annealing
50 h. Series 3 iron powders were milled for 100 h and then
annealed under argon atmosphere at 550 1C for 1 h.
Fig. 1 shows SEM micrographs for two different
Series 4 iron powders were milled for 50 h and annealed
powders, series 1 and 4. It is observed that the particle
under argon atmosphere at 550 1C for 1 h. The powders
size distribution for the milled powder without annealing
were milled again for 50 h and annealed at 550 1C for 1 h.
treatment is wider than the powder that had two steps
Table 1 lists different milling and annealing conditions
milling and annealing process.
used in the experiments. Iron powders were mixed in a
X-ray diffraction measurements were performed for all
mixer with continuous addition of 3 wt% epoxy resin and
powders. Fig. 2 presents the X-ray patterns for iron
hardener solution in acetone solvent. Powders and resin
powders listed in Table 1. It is clear that peak broadening is
were mixed in a mixer for 4 h at 80 1C to obtain a
lower for powders with annealing process. As some
parameters including particle size, powder composition,
Table 1
Powders prepared by different milling and annealing conditions
resin content and compaction pressure were kept constant,
the powders with high residual stress distortion show
Series Milling Annealing Milling Annealing higher broadening.
no. time (h) condition (1) time (h) condition Residual stresses in a material produce a distribution of
(1) (2) (2)
both tensile and compressive forces, which cause peak
Series 1 100 – – – broadening about the original position in the XRD
Series 2 50 550 1C/1 h/Argon 50 – patterns. Micro-stress in crystallites come from a number
Series 3 100 550 1C/1 h/Argon – – of sources such as vacancies, defects, shear planes, thermal
Series 4 50 550 1C/1 h/Argon 50 550 1C/1 h/
expansions and contractions. Whatever the cause of the
Argon
residual stress in a crystallite, the effect will cause a
ARTICLE IN PRESS
H. Shokrollahi, K. Janghorban / Journal of Magnetism and Magnetic Materials 317 (2007) 61–67 63

(110)
1000

Counts/s
(211)
500
(200) (220)

0
20 30 40 50 60 70 80 90 100 110 120 130
Position [°2 Theta]

1000

Counts/s
500

0
20 30 40 50 60 70 80 90 100 110 120 130
Position [°2 Theta]

1000
Counts/s

500

0
Fig. 1. SEM micrographs for two different powders: (A) series 1 and 20 30 40 50 60 70 80 90 100 110 120 130
(B) series 4. Position [°2 Theta]

distribution of d-values about the normal, unstrained, dhkl


value [9]. Table 2 lists the lattice strain for different samples 1000
which was concluded by Williamson–Hall model [9]. The
Counts/s

powders that have not experienced annealing treatment


have higher residual stresses and their diffraction profiles 500
broaden more (Fig. 2). The broadening in a peak is related
to the residual stress e, by bs ¼ 4e tan y [9], while crystallite
size broadening follows a 1/cos y rule which allows one to 0
separate these effects. 20 30 40 50 60 70 80 90 100 110 120 130
Fig. 3 depicts the specific resistivity of different samples. Position [°2 Theta]
The resistivity of the composite material strongly depends Fig. 2. X-ray diffraction patterns of four different powders: (A) Series 4,
on the amount of resin and defects such as porosity, point (B) Series 3, (C) Series 2 and (D) Series 1.
defects, residual stresses, distortions and dislocation
density. Annealing treatment can reduce these imperfec- Table 2
tions and release the residual stresses. Therefore, the Lattice strain for powders with different milling-annealing condition
samples which experienced annealing process show lower Series 1 Series 2 Series 3 Series 4
resistivity. The changes in temperature during two steps
annealing produced thermal stresses. Generates forces e (%) 0.75 0.6 0.45 0.2
causing the glissile threading dislocations to move; in
many cases such movement is toward another dislocation.
The two steps annealing process cause reduction in sessile stress-induced glissile dislocation glide and resulting
threading dislocation density by transformation from annihilation [10]. Another dislocation reduction mechan-
sessile to glissile threading dislocation and by thermal ism suggests that dislocation glide is induced by thermal
ARTICLE IN PRESS
64 H. Shokrollahi, K. Janghorban / Journal of Magnetism and Magnetic Materials 317 (2007) 61–67

130 3.2. Magnetic annealing


125 Series 4
120 Series 3 Fig. 6 presents SEM micrographs for different powders.
115 Series 2 The irregular nature of the iron powder particles is evident
ρ×104 (Ω.m)

110 Series 1 in Fig. 6A. The powder which was milled for 50 h has lower
105 particle size and particles shape have been become more
100 uniform (Fig. 6B).
95 Fig. 7 shows the specific resistivity for several samples.
90 At frequencies below 10 kHz, resistivity is almost constant
85
80
100 1000 10000 11.5 Series 4

Frequency (Hz) Series 3


10.5

ω0tan δ×10-3 (Ω/H)


Series 2
Fig. 3. Specific resistivity as a function of frequency (logarithmic level) at 9.5 Series 1
low frequencies.
8.5

7.2 7.5

7.1 Series 4 6.5


7 Series 3 5.5
6.9 Series 2 100 1000 10000
6.8 Frequency (Hz)
Series 1
6.7 Fig. 5. Magnetic loss as a function of frequency (logarithmic level) at low
μeff

frequencies.
6.6
6.5
6.4
6.3
6.2
6.1
100 2100 4100 6100 8100 10100
Frequency (Hz)

Fig. 4. Effective permeability as a function of frequency at low


frequencies.

stress imposed by the annealing cycle. The dislocation


annihilation mechanism described above can also occur
during dislocation glide toward a sink region [10]. Fig. 4
shows the effective permeability as a function of frequency.
Fig. 5 depicts the magnetic loss versus frequency. The
annealing of the milled powders for 50 and 100 h at 550 1C
during 1 h reduces the magnetic loss. The decrease of
magnetic loss can be due to residual strain reduction. When
an inductor core is exposed to a varying magnetic field,
losses originate in the core material. These losses can be
divided into three types depending on the physical back-
ground of the loss. The types are hysteresis, eddy current
and anomalous loss. At low-to-medium frequencies hyster-
esis losses dominate all others and total loss can be
expressed by that. The hysteresis loss is partly due to
stresses introduced in the material at compaction which
can impede domain wall movement [11]. Therefore to
reduce hysteresis in the iron-based composite, a stress
relieving low-temperature heat treatment or magnetic-field Fig. 6. SEM micrographs of iron powders: (A) unmilled powder and
annealing most often follows the compaction. (B) milled powder for 50 h.
ARTICLE IN PRESS
H. Shokrollahi, K. Janghorban / Journal of Magnetism and Magnetic Materials 317 (2007) 61–67 65

and at frequencies above 10 kHz, resistivity increases


noticeably. Resistivity is a function of particle size,
frequency, particle composition, amount of resin, internal
defects and density (porosity). In this study, all samples
had similar processing conditions, except post-compaction
or annealing treatment. Fig. 8 depicts a reproduction of the
high-frequency section of Fig. 7; the high-temperature
magnetic annealed sample has lower resistivity. At high
temperature (280 1C) contact points on the particle
surfaces, because of magnetic force increase, and further-
more, strong bonding between iron powder particles are
promoted because the epoxy resin is softened due to
heating and is pushed into particle interstices and residual
stresses and defects decrease. Fig. 9A shows the sample
without magnetic–thermal annealing. Fig. 9B illustrates at
high-temperature epoxy resin was softened and pushed into
interstices and at the same time, because of presence of
magnetic force, particles close to each other. The reasons
for reduction in total loss in the sample with magne-

1600
Without annealing
1400
Magnetic annealing
1200
low temperature-
magnetic annealing
ρ×104 (Ω.m)

1000
High temperature-
800 magnetic annealing
600
Fig. 9. SEM micrographs for the sample (A) without annealing treatment
400 (B) magnetic annealling at 0.5 T/280 1C for 300 s.

200
tic–thermal annealing at low frequencies can be related to
0 the reduction in hysteresis loss, residual stresses, distortion
100 1000 10000 100000
within particles and high uniformity. Behavior of magnets
Frequency (Hz)
is better characterized by considering them at low and high
Fig. 7. Specific resistivity as a function of frequency (logarithmic level) frequencies (o10kHz).
between 0.1 and 100 kHz. Fig. 10 shows the magnetic loss (o0 tan d) as a function
of frequency at low-frequency levels. Though the variation
of the loss is not remarkable at low frequencies, the
1400 Without annealing magnetic loss in the magnetic-field annealed state is smaller
Magnetic annealing
than that in the non-field annealed state. Also, this loss
1200
factor in the thermal–magnetic annealed state is smaller
Low temperature-
1000 magnetic annealing
than that in the magnetic annealed state. The loss variation
ρ×104 (Ω.m)

is different at higher frequencies 420 kHz as is discussed


High temperature-
800 magnetic annealing in the next two paragraphs.
When an inductor core is exposed to a varying magnetic
600 field, losses originate in the core material. These losses can
be divided into three types depending on the physical
400
background of the loss. The types are hysteresis, eddy
200 current and anomalous loss. At high frequencies, total loss
can be expressed by eddy current loss. At low-to-medium
0 frequencies hysteresis loss which strongly depends on
10000 100000
pinning sites, dominate all others and total loss can be
Frequency (Hz)
expressed by that.
Fig. 8. Specific resistivity as a function of frequency (logarithmic level), Fig. 11 depicts magnetic loss as a function of frequency
high frequencies (410 kHz). at frequencies higher than 10 kHz. It is clear that the
ARTICLE IN PRESS
66 H. Shokrollahi, K. Janghorban / Journal of Magnetism and Magnetic Materials 317 (2007) 61–67

10 7.5
Without annealing Without annealing
7.3
Magnetic annealing
9.5 Magnetic annealing
7.1 Low temperature-
Low temperature- magnetic annealing
9 magnetic annealing 6.9 High temperature-
magnetic annealing
High temperature- 6.7
8.5 magnetic annealing

μeff
6.5
ω0tan δ×10-3 (Ω/H)

6.3
8
6.1
7.5 5.9
5.7
7
5.5
100 2100 4100 6100 8100
6.5
Frequency (Hz)

6 Fig. 12. Effective permeability as a function of frequency at low


frequencies.

5.5
100 1000 10000 strongly depends on the material characteristic and is
Frequency (Hz) independent of material geometry. It is clear that the
Fig. 10. Magnetic loss as a function of frequency (logarithmic level) at low effective permeability in the magnetic-field annealed state is
frequencies (o10 kHz). larger than that in the non-field annealed state. Also this
parameter in the thermal magnetic annealed state is larger
than that in the magnetic field annealed state. The magnetic
140 moments are forced to orient along the magnetic field
Without annealing direction by an external magnetic field. For soft magnetic
120
Magnetic annealing
materials it is well known that a well-defined uniaxial
anisotropy can be obtained due to the atomic pair
Low temperaturemagnetic
100 annealing anisotropy induced by the magnetic field annealing [12].
ω0tan δ×10-3 (Ω/H)

High temperature-
magnetic annealing
80 4. Conclusions

60 In this paper, the effect of two steps and magnetic


annealing on the magnetic and electrical properties of iron
40 powder with high purity used in soft magnetic composites
was investigated. It was found that magnetic loss of
20 annealed powder was smaller than that of unannealed
powder after 100 h milling process. Annealing treatments
0 can eliminate residual stresses and some internal defects
10000 100000
and help to domain growth and domain wall movement.
Frequency (Hz)
Also, the magnetic loss of sample which had two steps
Fig. 11. Magnetic loss as a function of frequency at high frequencies milling and annealing process was smaller than the samples
(410 kHz). with one-step process. It was found that at low frequencies
(o10 kHz) the magnetic loss in the high-temperature
magnetic loss in the thermal magnetic annealed sample is magnetic annealed state is smaller than that in the low-
larger than the other samples. The reason for this behavior temperature magnetic, magnetic annealed and without
could be an effect of compaction of particles to each other annealing state. On the contrary, the magnetic loss in the
under the magnetic field. This compaction causes both a high-temperature magnetic annealed state is larger at high
trivial contact of some iron particles and limited thinning frequencies (410 kHz) than the others.
of insulator layer among other particles (Fig. 9B). A comparison of Figs. 5 and 10 suggests that two steps
Degradation of the surface insulation layer results in milling and annealing of the powders (series 4) decreased the
particle-to-particle contact and higher eddy current loss in magnetic loss similar to the high-temperature magnetic
the component. annealing at low frequencies o10 kHz. Figs. 4 and 12 also
Fig. 12 shows the effective permeability versus fre- show improvement of permeability for the two steps and high-
quency. Effective permeability is an important factor that temperature magnetic annealing compared to other conditions
ARTICLE IN PRESS
H. Shokrollahi, K. Janghorban / Journal of Magnetism and Magnetic Materials 317 (2007) 61–67 67

Acknowledgment [5] C. Ishihara, T. Ito, K. Asaka, Hitachi Powdered Metals Technical


Report 4 2005, pp. 30–33.
[6] Y.G. Guo, J.G. Zhu, J.J. Zhong, J. Magn. Magn. Mater. 302 (2006)
The authors express their attitude to the Fars Technical
14.
Research Commission for partial support. [7] L.A. Dobrazanski, M. Drak, J. Mater. Process. Technol. 157–158
(2004) 650.
[8] K. Asaka, C. Ishihara, Hitachi Powdered Metals Technical Report 4,
References 2005, pp. 3–9.
[9] E. Lifshin, X-ray Characterization of Materials, Wiley-VCH, 1999.
[1] E. Bayramli, O. Golgelioglu, H.B. Ertan, J. Mater. Process. Technol. [10] US Patent No. 6635110.
161 (2005) 83. [11] C.R. Hendricks, V.W.R. Amarakoon, D. Sullivan, Ceram. Bull. 70
[2] H. Shokrollahi, K. Janghorban, Mater. Sci. Eng. B 134 (2006) 41. (1991) 817.
[3] I.P. Gilbert, V. Moorthy, S.J. Bull, J.T. Evans, A.G. Jack, J. Magn. [12] D.S. Li, H. Garmestani, S-S Yan, M. Elkawni, M.B. Bacaltchuk, H.J.
Magn. Mater. 242 (2002) 232. Schneidert-Muntau, J.P. Liu, S. Saha, J.A. Barnard, J. Magn. Magn.
[4] A.G. Jack, B.C. Mecrow, P.G. Dickinson, In: Proceedings of IEEE Mater. 281 (2004) 272.
International Electric Machines and Drives Conference, Seattle, [13] D.A. Molodov, U. Czubayko, G. Gottstein, L.S. Shvindlerman,
USA, 1999. Acta. Mater. 46 (1998) 553.

You might also like