MRS Advances © 2016 Materials Research Society

DOI: 10.1557/adv.2016.29

Direct Observation of Reverse Magnetic Domain and Magnetic Domain Wall Motion in
Nd-Fe-B Magnet at High Temperature by Lorentz Microscop

Toshimasa Suzuki1, Koichi Kawahara1, Masaya Suzuki1, Kenta Takagi2 and Kimihiro Ozaki2
1
Materials Research and Development Laboratory, Japan Fine Ceramics Center (JFCC), Nagoya,
456-8587, Japan.
2
Inorganic Functional Materials Research Institute, National Institute of Advanced Industrial
Science and Technology (AIST), Nagoya, 463-8560, Japan.

ABSTRACT

We conducted the in-situ observations of the magnetic domain structure change in Nd-
Fe-B magnets at high temperature by transmission electron microscopy (TEM) / Lorentz
microscopy with applying an external magnetic field. Prior to observation, a thin foil was
magnetized by an external magnetic field of 2.0 T to almost saturation, then the magnetic domain
structures were observed by the Fresnel mode with in-situ heating. At 225oC, reverse magnetic
domains were found to generate in the thin foil sample without applying an external magnetic
field. When we applied a magnetic field on the same direction to the pre-magnetization direction
at 225oC, one magnetic domain wall was pinned by a grain boundary and the other magnetic
domain wall moved. As the results, the reverse magnetic domain shrank then annihilated. When
we cut the applied magnetic field, the reverse magnetic domain generated at almost the same
location. On the other hand, when we applied a magnetic field to the foils in the opposite
direction, the reverse domain started to grow, i.e., magnetic domain walls started to move. The
observation results of the shrink or growth of the reverse domain showed that the pinning effect
of grain boundary against domain wall motion would be different depending on the applied
magnetic field direction. Moreover, domain walls was observed to be pinned by grain boundaries
at elevated temperature, so that the coercivity of Nd-Fe-B magnet would occur by pinning
mechanism.

INTRODUCTION

Nd-Fe-B magnets which have excellent hard magnetic properties have extended the
applications in the field of motors for hybrid electric and electric vehicles. The permanent
magnet motors are often required to be used at elevated temperature as high as 200oC. It is well
known that the Nd-Fe-B magnets without dysprosium (Dy) show the decrease in coercivity at
elevated temperatures. In order to realize high coercivity of sintered Nd-Fe-B magnets even at
high temperature, various methods have been employed [1-3]. However, the coercivity of
developed Nd-Fe-B magnets has not reached to the coercivity of Nd-Fe-B with Dy [4]. Therefore,
in order to reduce deterioration of coercivity of Nd-Fe-B magnets at elevated temperatures, it is
important to observe generating sites of reverse magnetic domains and pinning sites against
magnetic domain walls motion in Nd-Fe-B magnets. The observations of magnetic domain
structure change in Nd-Fe-B magnets at room temperature by transmission electron microscopy
(TEM) have been reported in refs. 5 and 6. However, since the usage environment of magnet is
required around 200oC, the observation of magnetic domain structure at high temperatures would
be necessary. In this study, we conducted the in-situ observations of the magnetic domain

241
structure change in Nd-Fe-B magnets at high temperature by TEM / Lorentz microscopy with
applying an external magnetic field.

EXPERIMENT

Fine-grained sintered Nd-Fe-B magnets without Dy were prepared using

hydrogenation-disproportionation-desorption-recombination (HDDR) processed Nd-Fe-B
powder. Details of preparation of sintered magnets were described in ref. 7. The average grain
size of sintered Nd-Fe-B magnet was about 380 nm. A coercivity of the sintered Nd-Fe-B
magnet is 1.2 T. Prior to preparation of thin foils suitable for TEM observations, the sintered Nd-
Fe-B magnet was thermally demagnetized at 350oC in Ar atmosphere.
Thin foils with dimensions of ca. 10 ȝm x 10 ȝm x 100 nm were prepared by a focused
ion beam (FIB) thinning method (Hitachi FB-2100). The thin foils were supported on Mo grids.
In order to observe the thermal demagnetization process in TEM, the prepared foils were
magnetized using an electromagnet with applying an external magnetic field of 2.0 T. Then, foils
were subjected to TEM observation. A Hitachi HF-3300EH was used in this study with
acceleration voltage of 300 kV. In-situ observation of the magnetic domain wall motion at high
temperatures was carried out using a double-tilt heating holder at the temperature range from
room temperature to 225oC.
Magnetic domain walls were observed by the Lorentz TEM microscopy with Fresnel
mode at the same area. The electron microscope used in this study is specially designed suitable
for Lorenz observation; application of an external magnetic field is designed for using the
condenser lens. In-situ observation of domain wall motion was conducted by applying a
magnetic field from condenser lens in TEM. Hereafter, positive and negative values of the
applied magnetic field mean the same and the opposite direction as the initial magnetizing
direction, respectively. The external magnetic field was applied up to 100 mT in TEM, and the
direction of the applied magnetic field was parallel to the incident beam. During observation, the
sample was tilted by 30o, so that the effective magnetic field, that was the magnetic field
component parallel to the easy magnetization direction, was the range of -30.9 to 30.9 mT.

RESULTS and DISCUSSION

Prior to observation, a thin foil was magnetized by an external magnetic field of 2.0 T to
almost saturation, then the magnetic domain structures were observed by the Fresnel mode with
in-situ heating. The reverse magnetic domains were found to generate at 225oC in the thin foil
sample without applying an external magnetic field. As shown in Figure1 (a) by the arrows 1 and
2, the dark and bright line contrasts of magnetic domain walls appeared in the observed area at
225oC (Hereafter, the domain walls indicated by the arrows 1 and 2 are referred as domain wall 1
and domain wall 2, respectively in all figures). The magnetic domain walls were observed to
locate along grain boundaries, indicating that grain boundaries act as pinning sites for domain
wall motion.
Figure 1 shows the magnetic domain walls motion with applied magnetic field on the
same direction to the pre-magnetization direction at 225oC. Although the magnetic field
increased up to 11 mT, the magnetic domain walls motion was not observed (Fig. 1(a)-(f)).
When the magnetic field was increased up to 13.2 mT, the domain wall 1 moved toward the right
side of the observation area (Fig. 1(g)). When the magnetic field was increased up to 15.4 mT, a

242
part of domain wall 1 moved (Fig. 1(h)). As further increasing the magnetic field up to 17.7 mT,
the magnetic domain walls annihilated in the observation area (Fig. 1(i)). In Fig. 1, the motion of
the domain wall 2 was not observed with increasing magnetic fields. These observation results
indicated that the pinning force for the domain wall 1 is weaker than that for domain wall 2. Next,
we applied magnetic field to the opposite direction with pre-magnetization direction and
magnetic domain wall motion was observed.

(a) (b) (c)

1 1 1

2
2
H 2

(d) (e) (f)

1 1 1

2 2 2

(g) (h) (i)

1 1
2 2

2 ȝm
Figure 1 Lorentz microscope images observed at 225oC with different strength of the applied
magnetic field: (a) 0 mT, (b) 2.2 mT, (c) 4.4 mT, (d) 6.6 mT, (e) 8.8 mT, (f) 11.0 mT, (g) 13.2
mT, (h) 15.4 mT, (i) 17.7 mT. The arrows 1 and 2 show the magnetic domain walls. The
arrow H represents the direction of the applied magnetic field.

Figure 2 shows the magnetic domain walls motion with the opposite direction of applied
magnetic field to the pre-magnetization direction at 225 oC. When we cut the applying magnetic
field after observations shown in Fig.1, the reverse magnetic domain generated at almost the
same location (see Fig. 1(a) and Fig. 2(a)). Therefore, the region where the reverse domain
generated (the region between the domain walls 1 and 2) would be a preferential generation site
for reverse magnetic domain. The magnetic domain wall motion was not observed even though

243
we increased magnetic field from 0 to -4.4 mT. As shown in Fig.2 (d)-(g), when the magnetic
field increased from -6.6 to -13.2 mT, the reverse domain started to grow, i.e., magnetic domain
walls started to move. As increasing the magnetic field up to -6.6 mT, a part of domain wall 2 (as
shown in Fig. 2 (d) by the arrows B) was unpinned and jumped to the neighboring grain
boundaries. However, the pinned part of domain wall 2 (as shown in Fig. 2 (d) by the arrow A)
did not move from pining site of the grain boundary. When the magnetic field increased up to 8.8
mT, the pinned part of the domain wall 2 (shown in Fig. 2 (d) by the arrows A) was unpinned.
Then, the domain wall jumped to the next pinning site. As further increasing the magnetic field,
the domain wall 2 moved toward the right side of the observation area. As further increasing the
magnetic field up to -15.4 mT, the magnetic domain wall 2 went out of the observation area. In
Fig. 2, the motion of domain wall 1 was not observed with increasing magnetic fields. These
observation results indicated that the pinning force for the domain wall 2 is weaker than that for
domain wall 1. These results were different from the results shown in Fig. 1 as mentioned above.
The obtained results from the observations shown in Figs. 1 and 2 suggested that the pinning
force of grain boundaries against domain walls motion would differ depending on the direction
of the domain wall motion. Therefore, we applied magnetic field to the same direction with pre-
magnetization direction and magnetic domain wall motion were observed again for confirmation
of the repeatability.

(d) A
(a) 1
H (b) 1 (c) 1

2 1
2 2
B

(e) (f) (g) (h)

1 1 1 1
2
2
2
2 ȝm
Figure 2 Lorentz microscope images observed at 225RC with different strength of the applied
magnetic field: (a) 0 mT, (b) -2.2 mT, (c) -4.4 mT, (d) -6.6 mT, (e) -8.8 mT, (f) -11.0 mT, (g) -
13.2 mT, (h) -15.4 mT. The arrows 1 and 2 show the magnetic domain walls. The arrows A and
B in Fig.2 (d) show the pinned and unpinned parts of domain wall. The arrow H represents the
direction of the applied magnetic field.

When we cut the applied magnetic field, unlike in Fig. 2(a), the magnetic domain walls
showed a different configuration with that shown in Figs 1(a) and 2(a). When we increased the
magnetic field up to 13.2 mT, the domain structure in the observation area changed to almost the

244
same configuration as shown in Figs. 1(a) and 2(a) after the relatively complicated motion of
domain walls. As shown in Fig.3, the magnetic domain walls were pinned to the same grain
boundary as shown in Fig.1. The domain wall 1 moved toward the right side of the observation
area with increasing the magnetic field from 13.2 to 22.1mT (Fig. 3(a)-(g)). Eventually, the
domain walls annihilated when the magnetic field is increased to 24.3mT (Fig. 3(h)). Those
observation results were the same as those obtained from the observations shown in Fig.1,
indicating that the pinning force for domain wall 1 was weaker than that for domain wall 2.

(a) (b) (c)

1 1 1

2
2
H 2

(d) (e) (f)

2 2
1 1

2 ȝm
Figure 3 Lorentz microscope images observed at 225oC with different strength of the applied
magnetic field: (a) 13.2 mT, (b) 15.4 mT, (c) 17.7 mT, (d) 19.9 mT, (e) 22.1 mT, (f) 24.3 mT.
The arrows 1 and 2 show the magnetic domain walls. The arrow H represents the direction of
the applied magnetic field.

The domain wall motion was observed at 225oC in Nd-Fe-B magnets. The domain wall
motion easily occurred at 225oC under the applied magnetic field of less than 20 mT. During
observations, some grain boundaries act as pinning site for domain wall motion. These domain
wall motion have been reported in ref. 6 and 8. Moreover, when we applied the magnetic field to
the same direction with pre-magnetization direction (Figs. 1 and 3), domain wall 2 was pinned
by the grain boundary and domain wall 1 moved. On the hand, when magnetic field applied the
opposite direction with pre-magnetization direction (Fig. 2), domain wall 2 moved and domain
wall 1 was pinned by the grain boundary. Thus, the pinning effect of grain boundaries would be
different depending on the applied magnetic field direction.
Generally, Nd-Fe-B magnet has been thought the nucleation type. Recently, however, it
has been reported that the grain boundaries act as an effective pinning sites for domain wall
motion in Nd-Fe-B magnets at room temperature [6, 9-11]. In this study, observations of domain
wall motion indicated that grain boundaries acted as an effective pinning sites for domain wall
motion at high temperature, so that the coercivity of Nd-Fe-B magnet would occur by pinning
mechanism at high temperature.

245
SUMMARY

The magnetic domain structure change in fine-grained sintered Nd-Fe-B magnets without
Dy was observed at high temperature by transmission electron microscopy (TEM) / Lorentz
microscopy with applying an external magnetic field. Obtained results are summarized as
follows;

(1) The generation of the reverse magnetic domain and the domain wall motion in Nd-Fe-B
magnets were successfully observed at high temperature of 225oC.
(2) The magnetic domain wall motion easily occurred by applying the magnetic field of less than
20 mT at 225oC, and some grain boundaries act as pinning site against domain wall motion.
(3) The pinning effect was suggested to be different depending on the direction of the applied
magnetic field.
(4) The coercivity of Nd-Fe-B magnet would occur by pinning mechanism at high temperature.

ACKNOWLEDGMENTS

This work is based on results obtained from the future pioneering program "Development
of magnetic material technology for high-efficiency motors" commissioned by the New Energy
and Industrial Technology Development Organization (NEDO).

REFERENCES

1. 1. R. Ramesh, G. Thomas, and B. M. Ma, J. Appl. Phys. 64, 6417 (1988).

2. H. Sepehri-Amin, T. Ohkubo, T. Nishiuchi, S. Hirosawa, and K. Hono, Scr.Mater. 63, 1124
(2010).
3. T. Akiya, J. Liu, H. Sepehri-Amin, T. Ohkubo, K. Hioki, A. Hattori, K.Hono, J. Appl. Phys.
115, 17A766 (2014).
4. K.Hono, H. Sepehri-Amin, Scr.Mater. 67, 530 (2012).
5. H. S. Park, Y.-G. Park, Y. Gao, D. Shondo and M. Inoue, J. Appl. Phys. 97, 033908 (2005).
6. W.F.Li, T. Ohkubo, K. Hono, T. Nishiuchi, and S. Hirosawa, J. Appl. Phys. 105, 07A706
(2009).
7. K.Takagi, M. Akada, R.Soda, K. Ozaki, J. Magn. Magn. Mater. 393, 461 (2015).
8. T. Suzuki, K. Kawahara, H. Tanaka and K. Ozaki in State-of-the-Art Developments in
Materials Characterization, edited by R. Barabash , L.G. Benning , A. Genc , Y. Kim , A.
Lereu , D. Li , U. Lienert , K.D. Liss , M. Ohnuma , O. Ovchinnikova , A. Passian , J.D.
Rimer , L. Tetard , T. Thundat , R. Zenobi and V. Zorba, (Mater. Res. Soc. Symp. Proc. 1754,
2015) opl.2015. 585.
9. W. F. Li, T. Ohkubo, K. Hono, T. Niniushi, and S. Hirosawa, Appl, Phys. Lett. 93, 052505
(2008).
10. Y. Matsuura, N. Kitai, R. Ishii, T. Natsumeda, J. Hoshijima, J. Magn. Magn. Mater. 336, 88
(2013).
11. Y. Matsuura, R. Ishii, J. Magn. Magn. Mater. 346, 113 (2013).

246