Preparation of Iron Fine Particles Coated with Boron Nitride Layers

Materials Transactions, Vol. 45, No. 9 (2004) pp. 2941 to 2944
#2004 The Japan Institute of Metals
RAPID PUBLICATION
Preparation of Iron Fine Particles Coated with Boron Nitride Layers
Hisato Tokoro1 , Shigeo Fujii1 , Takeo Oku2 , Takashi Segi3 and Saburo Nasu3
1
Hitachi Metals, Ltd., Advanced Electronics Research Laboratory, Kumagaya 360-0843, Japan
Osaka University, Nanoscience and Nanotechnology Center, Institute of Scientific and Industrial Research,
Ibaraki 567-0047, Japan
3
Osaka University, Divison of Materials Physics, Department of Materials Engineering Science,
Graduate School of Engineering Science, Toyonaka 560-8531, Japan
2
Boron (B) powders mixed with hematite or iron (Fe) nanoparticles were annealed at 1573 K in nitrogen atmosphere. X-ray diffraction
measurement shows that mixture of B and hematite changed to boron nitride (BN) and metallic Fe after the annealing without the formation of
FeB compounds like FeB and Fe2 B. FeB compound was generated from a mixture of B and Fe. High-resolution transmission electron
microscopy shows that Fe particles were 300 nm in diameter and they were coated with BN layers. Mössbauer spectroscopy revealed a
reduction of hematite to Fe. Synthesized Fe fine particles have an oxidation resistance up to 673 K.
(Received June 14, 2004; Accepted July 20, 2004)
Keywords: iron fine particles, boron nitride layers, catalyst for boron nitride formation
1.
Introduction
Magnetic nanoparticles can be utilized for various applications such as high density magnetic recording media,1)
magnetic fluids,2) magnetic carrier in clinical cure,3) and
other novel magnetic devices. Metal nanoparticles such as Fe
and Fe-based alloys are expected to be a good candidate for
these applications because of their high magnetizations
compared to iron oxide. Unfortunately those nanoparticles
are easily oxidized in air, and the oxidation cause deterioration of their magnetizations.
In order to solve this problem, coating techniques have
been studied. An arc-discharge method was proposed to coat
metal nanoparticles with carbon4–6) or boron nitride (BN)7–9)
layers. This method is popular for coating, but it is hard to
treat a large quantity of nanoparticles due to the low
production yield. Fe nanoparticles coated with iron oxide
(-Fe2 O3 ) layers were synthesized by exposing Fe nanoparticles in a nitrogen/air atmosphere.10) Though the thickness of oxide layers are controllable by slight oxidization,
they cause the deterioration of the magnetic properties.
Kitahara et al. have developed a simple method for
production of a large quantity of cobalt (Co) nanoparticles
encapsulated in BN nanocages,11) which were synthesized by
annealing a mixture of Co nanoparticles and B powders at
1073 K in H2 /NH3 atmosphere with an ordinary furnace. In
our previous works, we have successfully fabricated Fe
nanoparticles coated with BN layers by annealing a mixture
of hematite (-Fe2 O3 ) nanoparticles and B powders at
1373 K in a nitrogen atmosphere.12,13) In these methods, the
metal nanoparticles behaved like a catalyst for BN formation.
The -Fe2 O3 nanoparticles act as a catalyst for the
formation of BN nanotubes.14) Metal Fe can also play an
active role in the growth of BN fibres.15) However we don’t
know whether the effective catalyst for the BN layers is iron
oxide or metal iron. The purpose of the present work is to
understand the mechanism of BN formation by using Fe2 O3 or Fe nanoparticles as a catalyst. In our previous
work,12) the differential thermal analysis showed that BN was
generated from B and -Fe2 O3 above 1473 K. In order to
generate BN fully, the mixtures were annealed at 1573 K in
the present work.
2.
Experimental
Commercial B powders, -Fe2 O3 and Fe nanoparticles
were used as starting materials. Their diameters were 27 mm,
30 nm and 20 nm respectively. These materials were mixed
well each other by a V-mixer. The weight ratio of B to Fe2 O3 or to Fe was 1:1. Samples were prepared by annealing
the mixtures at 1573 K for 2 hours in a nitrogen atmosphere.
The sample obtained from mixture of B and -Fe2 O3 was
denoted as sample I, and the sample obtained from mixture of
B and Fe was denoted as sample II.
X-ray diffraction measurements were employed to detect
the produced phases in the samples with a Cu-K radiation
under an applied power of 50 kV and 250 mA. Microstructure
of the samples was observed by a high-resolution electron
microscope (HRTEM) equipped with an energy dispersive
X-ray analysis (EDX), by which compositions of particles
were analyzed. Mössbauer spectroscopy with a 57 Co source
was conducted with a conventional constant-acceleration
method at room temperature in order to confirm the existence
of iron oxides and metalic Fe. Magnetic properties were
measured by a vibrating sample magnetometer (VSM) at
room temperature under a field of 1.6 MA/m. Magnetization
at 1.6 MA/m was called ‘‘saturation magnetization’’ in the
present work. Oxidation resistance of the sample was
measured as changes of saturation magnetizations after
annealing at 300–873 K for 1 hour in air.
3.
Results
Figures 1(a) and (b) show X-ray diffraction patterns of the
sample I and sample II, respectively. As shown in Fig. 1(a),
all peaks are assigned to -Fe and hexagonal or rhombohedral BN (h- or r-BN), and no peaks of starting materials are
detected. It means that -Fe2 O3 was reduced to Fe, and B
H. Tokoro, S. Fujii, T. Oku, T. Segi and S. Nasu
Intensity (a. u.)
2942
(a)
h-BN
r-BN
α-Fe
(b)
B
FeB
20°
30°
40°
50°
60°
70°
100 nm
80°
degrees, 2 θ
Fig. 1
(a)
(b)
X-ray diffraction patterns of the sample I (a) and II (b).
changed into BN by nitridation of B during annealing. In Fig.
1(b), strong peaks are assigned to FeB compound, and weak
peaks are corresponding to B, which is a starting material.
Annealing a mixture of B powders and metallic Fe nanoparticles usually forms the FeB compound, and BN was
hardly formed under a nitrogen atmosphere. This result
suggests that -Fe2 O3 nanoparticles behaved like a catalyst
for BN formation. However, Fe nanoparticles didn’t behave
like the catalyst in the present work, and FeB compound was
only formed.
Figure 2 shows HRTEM images of the sample I. A particle
with dark contrast can be seen in Fig. 2(a), the size is
300 nm in diameter. An EDX analysis confirms that the
particle consists of Fe. A detailed morphology on the surface
of the Fe particle is shown in Fig. 2(b). It reveals that the Fe
particle has a shell structure, in which nanometer layers with
the thickness of 10 nm encapsulate the Fe core. Spacing
between lattice fringes in the layers was measured as
0.33 nm, and it is equal to (002) spacing of h-BN. Therefore
it turns out that Fe fine particles in the sample I are coated
with h-BN layers.
Figure 3 shows a Mössbauer spectrum about the sample I.
The spectrum has sextet, and the hyperfine field was
evaluated as 32.4 T, which is typical value for -Fe. This
proves that -Fe2 O3 nanoparticles are reduced to -Fe almost
completely after the annealing.
VSM measurement shows that saturation magnetization of
the sample was 47 Am2 /kg and the coercivity was 2.0 kA/m.
Though the magnetic properties would come from the
reduced -Fe, the saturation magnetization is 22% of bulk
-Fe16) because the sample I include non-magnetic BN. The
coercivity is low compared to other Fe or Co nanoparticles,
of which coercivities were around 20 kA/m.17–19) Accordingly the sample I exhibits excellent soft magnetic properties
compared to other magnetic nanoparticles.
In order to check oxidation resistance of the Fe fine
(002)BN
Fe
1 nm
Fig. 2 Morphology of Fe fine particles in sample I; (a) TEM image of a Fe
fine particle, (b) magnified image on the surface of the Fe fine particle.
Fig. 3
Mössbauer spectrum of the sample I measured at room temperature.
particles coated with BN layers, the sample I was annealed in
air. Saturation magnetizations (= MT ) after annealing at each
temperature for 1 h were measured at room temperature, and
they were normalized by the saturation magnetization (= M0 )
Degradation of magnetization
Preparation of Iron Fine Particles Coated with Boron Nitride Layers
2943
Table 1 Gibbs’s free energy (G) for chemical reactions represented by
eqs. (1)–(3) at 1573 K. G at a temperature (T) was calculated following
an equation, G ¼ H TS. H and S mean enthalpy and entropy,
respectively. They are calculated from thermodynamic data, that is
standard enthalpy, standard entropy and heat capacity, adopted from
Ref. 23.
1.0
0.8
0.6
(kJ)
0.4
0.2
G (1573 K)
Formula (1)
Formula (2)
Formula (3)
666
146
219
0.0
273
473
673
873
Annealing temperature, T/ K
Fig. 4 Degradation of magnetization of samples after annealing at 300–
873 K in air. Black circles ( ) represent a result of sample I, and white
circles ( ) that of non-coated Fe nanoparticles.
was scarcely generated through the nitridation of the FeB
particles.
Chemical reactions between B and -Fe2 O3 or Fe are
given by following formulas,
Fe2 O3 þ 4B þ N2 ! 2FeB þ B2 O3 þ N2
! 2Fe þ 2BN þ B2 O3
before the annealing. Figure 4 shows degradation of magnetization (= MT /M0 ) as a function of the annealing temperature. Fe nanoparticles of the starting materials are used as
control, and the data are represented by white circle ( ) in
Fig. 4. The MT /M0 of the control sample begins to decrease
at 473 K. On the other hand, that of the sample I, which is
represented by black circle ( ) in Fig. 4, is kept at constant
value up to 673 K, and decrease abruptly above 773 K. Since
the degradation of the magnetization is caused by oxidation
of Fe fine particles, this result proves that the BN layers
protect the Fe fine particles against the oxidation.
4.
Discussions
Huo and the coworkers mentioned that BN were generated
by nitridation of FeB nanoparticles at 1373 K in NH3 /N2
atmosphere, following the Vapor-Liquid-Solid (VLS) mechanism.20,21) Huo suggested that the FeB nanoparticles should
be in a molten or quasi-liquid state at 1373 K, which is lower
temperature than the melting point of the bulk FeB, so that N
atoms from the atmosphere easily diffuse into the FeB
‘‘catalyst’’, and when they become supersaturated, BN is
precipitated in a solid state. As Fe-B compounds were
generated in the process of the reaction between B and Fe2 O3 ,12) it is considered that BN in the sample I was
generated through the decomposition of Fe-B compounds,
following the VLS mechanism. In heating process, B
powders probably reduced -Fe2 O3 nanoparticles to Fe,
and excess B immediately react with the reduced Fe,
resulting in a generation of FeB fine particles separated by
a produced boron oxide. Then, N atoms are taken in the FeB
fine particles, which would be in a molten state, and BN
layers are precipitated on the surface when the N atoms are
supersaturated in the FeB particles during the annealing. The
X-ray pattern (Fig. 1(a)) and the Mössbauer spectrum (Fig. 3)
show that FeB compounds have been completely decomposed into Fe and BN after the annealing. In the case of the
sample II, B reacted with Fe nanoparticles, resulting in the
FeB compound. It is considered that the FeB particles were
not kept fine, but coarsely agglomerated each other. The
melting point probably was not depressed, so that it is
difficult for N atoms to diffuse into the FeB. Therefore, BN
2Fe þ 2B ! 2FeB
ð1Þ
ð2Þ
The B2 O3 , which become liquid state above 850 K,22)
would prevent FeB fine particles from agglomerating each
other, and help the nitridation of the FeB fine particles. If BN
were generated from Fe and B, the chemical reaction is given
by following formula,
2Fe þ 2B þ N2 ! 2FeB þ N2 ! 2Fe þ 2BN
ð3Þ
In order to confirm the chemical reactions given by the
formulas (1)–(3), Gibbs free energies (G) for each reaction
were calculated. Enthalpies (H) and entropies (S) at 1573 K
about each element and compound were calculated from
standard enthalpies, standard entropies and heat capacities
adopted from the Ref. 23, and then G at 1573 K were
obtained from the formula; G ¼ H TS. The G difference between the final products generations and that of
starting materials about each formula represents the G for
each reaction (see Table 1). Since G for each formula are
negative, the chemical reactions given by the formula (1)–(3)
are reasonable at 1573 K. However, the reaction given by the
formula (3) didn’t proceed in the present work, being
attributed to the agglomeration of the FeB particles.
The reaction given by the formula (1) could proceed easily
compared to the reactions given by other formula, because
the G is minimum. This means that the BN grow stably
when B powders are annealed with -Fe2 O3 nanoparticles in
a nitrogen atmosphere. Since -Fe2 O3 nanopartilces stabilize
the BN formation, it is considered that they behave like a
catalyst for the formation of BN layers.
As shown in Fig. 4, the BN layers protected the Fe core
particles against the oxidation in air, because the degradation
of magnetizations were kept at 1 up to 673 K. Accordingly,
the BN layers intercept the Fe fine particles from oxygen
completely. The decrease of the magnetization above 673 K
means that the BN layers were broken by the oxidization, so
that the Fe fine particles were abruptly oxidized.
5.
Conclusion
In conclusion, Fe fine particles coated with BN nanolayers
have been successfully prepared by annealing a mixture of B
2944
H. Tokoro, S. Fujii, T. Oku, T. Segi and S. Nasu
and -Fe2 O3 at 1573 K in a nitrogen atmosphere. The
-Fe2 O3 nanoparticles were completely reduced to a-Fe fine
particles with 300 nm in diameter. The Fe fine particles
were expected as an excellent soft magnetic material by the
virtue of the low coercivity and resistance oxidation. The
method for synthesis of the Fe fine particles were really a
simple because of using an ordinary furnace and harmless
atmosphere, and it is good candidate to provide the Fe fine
particles coated with BN nanolayers for mass-production.
Considering the Gibbs free energy about each reaction,
-Fe2 O3 nanoparticles successfully behave like a catalyst for
the BN formation.
REFERENCES
1) F. E. Kruis, H. Fissan and A. Peled: J. Aerosol Sci. 29 (1998) 511–535.
2) J. Ding, W. F. Miao, P. G. McCormick and R. Street: Appl. Phys. Lett.
67 (1995) 3804–3806.
3) Q. Liu, Z. Xu, J. A. Finch and R. Egerton: Chem. Mater. 10 (1998)
3936–3940.
4) Y. Yoshida, S. Shida, T. Ohsuna and N. Shiraga: J. Appl. Phys. 76
(1994) 4533–4539.
5) X. L. Dong, Z. D. Zhang, S. R. Jin and B. K. Kim: J. Appl. Phys. 86
(1999) 6701–6709.
6) Z. D. Zhang, J. G. Zheng, I. Skorvanek, G. H. Wen, J. Kovac, F. W.
Wang, J. L. Yu, Z. J. Li, X. L. Dong, S. R. Jin, W. Liu and X. X. Zhang:
J. Phys.: Condens. Matter. 13 (2001) 1921–1929.
7) T. Hirano, T. Oku and K. Suganuma: Diamond and Related Mater. 9
(2000) 476–479.
8) M. Kuno, T. Oku and K. Suganuma: Scr. Mater. 44 (2001) 1583–1586.
9) T. Oku, M. Kuno, H. Kitahara and I. Narita: Int. J. Inorg. Mater. 3
(2001) 597–612.
10) K. K. Fung, B. Qin and X. X. Zhang: Mater. Sci. Eng. A 286 (2000)
135–138.
11) H. Kitahara, T. Oku, T. Hirano and K. Suganuma: Diamond and
Related Mater. 10 (2001) 1210–1213.
12) H. Tokoro, S. Fujii and T. Oku: IEEE Trans. Magn. 39 (2003) 2761–
2763.
13) H. Tokoro, S. Fujii and T. Oku: J. Mater. Chem. 14 (2004) 253–257.
14) C. C. Tang, M. Chapelle, P. Li, Y. M. Liu, H. Y. Dang and S. S. Fan:
Chem. Phys. Lett. 342 (2001) 492–496.
15) L. Bourgeois, Y. Bando and T. Sato: J. Phys. D: Appl. Phys. 33 (2000)
1902–1908.
16) National Astronomical Observatory: Chronological Scientific Tables
2000, (Maruzen Co. Ltd., Tokyo, 1999) pp. 491.
17) T. Hayashi, S. Hirono, M. Tomita and S. Umemura: Nature 381 (1996)
772–774.
18) J. J. Host, J. A. Block, K. Parvin, V. P. Dravid, J. L. Alpers, T. Sezen
and R. LaDuca: J. Appl. Phys. 83 (1998) 793–801.
19) X. X. Zhang, G. H. Wen, S. Huang, L. Dai, R. Gao and Z. L. Wang:
J. Magn. Magn. Mater. 231 (2001) 9–12.
20) K. F. Huo, Z. Hu, J. J. Fu, H. Xu, X. Z. Wang, Y. Chen and Y. N. Lu:
J. Phys. Chem. B 107 (2003) 11316–11320.
21) K. F. Huo, Z. Hu, F. Chen, J. J. Fu, Y. Chen, B. H. Liu, J. Ding, Z. L.
Dong and T. White: Appl. Phys. Lett. 80 (2002) 3611–3613.
22) Tables of Chemical Constants for Laboratory Use, 2nd ed, (Hirokawa
Publishing Co., Tokyo, Japan, 1981) pp. 17.
23) D. R. Lide: CRC Handbook of Chemistry and Physics, 73rd ed, (CRC
Press, USA, 1992), pp. 5–9, pp. 5–17, pp. 5–18.