- Email: [email protected]
Microwave absorption properties of FeSi flaky particles prepared via a ball-milling process
Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Contents lists available at ScienceDirect
Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm
Microwave absorption properties of FeSi flaky particles prepared via a ball-milling process Chao Liu a, Yong Yuan b, Jian-tang Jiang a, Yuan-xun Gong c, Liang Zhen a,d,n a
School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China Precision Machinery Research Institute of Shanghai Space Flight Academy, Shanghai 201600, China c Aerospace Research Institute of Special Material and Processing Technology, Beijing 100074, China d MOE Key Laboratory of Micro-system and Micro-structures Manufacturing, Harbin Institute of Technology, Harbin 150080, China b
art ic l e i nf o
a b s t r a c t
Article history: Received 23 January 2015 Received in revised form 26 May 2015 Accepted 27 June 2015 Available online 8 July 2015
Flaky FeSi alloy particles with different aspect ratio were produced via ball-milling and a subsequent annealing. The microstructure and the morphology of the particles were examined by XRD and SEM. The dc resistivity, the static magnetization properties and electromagnetic properties were measured. Particles with high aspect ratio were found possess high permittivity and permeability. On the other hand, the variation of grain size and defects density was found influence the permittivity and permeability. High specific area was believed contribute to the intense dielectric loss and the high shape magnetic anisotropy lead to high permeability in the target band. Increased electromagnetic parameters compel the absorption peak’s shift to lower frequency. Coating using flaky FeSi particles milled for 12 h as fillers presented a reflection loss of 10 dB at 2 GHz and a matching thickness of 1.88 mm. The flaky FeSi alloy particles prepared through ball-milling and annealing can be promising candidates for EMA application at 1–4 GHz band. & 2015 Elsevier B.V. All rights reserved.
Keywords: FeSi particles Ball-milling Shape Microstructure Electromagnetic properties
1. Introduction With the fast development of wireless communication, the problem of electromagnetic interference (EI) and electromagnetic pollution (EP) has become more and more serious, especially in the 1–4 GHz band at which most communication systems work [1]. To absorb the harmful electromagnetic wave is an effective way to eliminate EI and EP problems, which basically rely on the application of electromagnetic wave absorbing (EMA) materials. Ferromagnetic metal/alloy powder of micro scale is the most promising candidate for EMA applications in the 1–4 GHz band since this series of materials possesses a unique potency of presenting a combination of high permeability and high permittivity [2]. To exert the potency of this series of materials is then a basic step to design high-performance EMA materials. High and commensurate electromagnetic (EM) properties (the complex permeability and the complex permittivity) ensure small matching thickness and good impedance matching [3,4], which is crucial for obtaining high EMA performance in 1–4 GHz band. A lot of researches have then been carried out to obtain high EM n Corresponding author at: School of Material Science and Engineering, Harbin Institute of Technology, Harbin 150001, China. Fax: þ 86 451 8641 3922. E-mail address: [email protected] (L. Zhen).
http://dx.doi.org/10.1016/j.jmmm.2015.06.085 0304-8853/& 2015 Elsevier B.V. All rights reserved.
properties. Among these efforts, particles with flaky shape and thin thickness have attracted intense attention, since they provide a possibility to obtain high permeability in 1–4 GHz band. The Snoek’s limit demonstrates [5,6] that it is essentially difficult to increases the permeability and the resonance frequency (fr) simultaneously, revealing the obstacle to achieve high permeability in GHz band. This limit, however, could be exceeded by introducing easy-plane anisotropy via using flaky particles [7–10]. Additionally, thin thickness is believed helpful to suppress the negative effects from the eddy current in metal or alloys particles [11, 12]. The eddy current induces reversed magnetization in particles and also reduces the effective volume due to the skin effect, which then leads to deteriorated permeability. Han et al. [13] flattened spherical iron particles into fine flakes and observed that particles with aspect ratio (AR) of 5 presented much higher permeability than that of spherical ones. A minimum reflection loss (RL) of 26 dB was observed at 3.5 GHz in a thin (2.0 mm) coating using flaky iron particles as fillers. Yang [14] prepared flaky Fe16Ni82Mo2 particles through a similar method and wherefrom achieved increased permeability and permittivity. The corresponding coating with a thickness of 2 mm presented a RL of 20.4 dB at 13.0 GHz. Liu et al. [15] prepared spherical FeSi alloy powders of various size by gas-atomization method and found that both the permeability and the permittivity increased as the particle size decreased. Among these studies, ball-milling was
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
found an efficient method to prepare flaky metal/alloy particles and improved EMA performances were wherefrom obtained. Most attention has been focused on the influence related to particles’ shape/size in the previous research, the evolution in microstructure characteristics including the grain size, the defects density and so on, however has not been investigated systematically. In this paper, flaky FeSi alloy particles were produced by ball-milling commercial FeSi alloy powders in large quantity to examine the feasibility. Post-milling annealing was thereafter performed to tailor the microstructure of the powder. The morphology and microstructure of the particles were examined and their influence on EM properties and the EMA performance were investigated.
2. Experimental details Commercially available gas-atomized FeSi alloy powder with mesh -400 was used as the raw material of the current study. Composition analysis indicated that the starting powder contain 6.31 wt% of Si, 0.07 wt% of Cr and allowance of Fe. The raw powder, certain content anhydrous ethanol and GCr15 steel balls were added into jars of 250 ml and then the ball-milling process was conducted in a planetary ball-mill at a rotation rate of 500 rpm for up to 20 h. Through the study, the ball–powder–ethanol ratio was set at 5:1:1. After the ball-milling completed, the product was collected, washed for three times with anhydrous ethanol and then dried for 8 h at 50 ºC in an oven. The product milled for 20 h was labeled ON500r20h for short and samples of other conditions were labeled in the same way. Some as-milled powder (ON500r20h) was transferred into annealing process. The power was heated in a furnace in hydrogen atmosphere at 800 ºC, hold for 2 h and then let cooled to room temperature naturally. The annealed powder was labeled as H500r20h. Phase structure of the particles, raw, milled and mill-annealed, was examined on a PANalytical X’Pert PRO diffractometer (XRD) using Cu Kα radiation. The grain size of the particles was calculated according to Scherer formula
D = Kλ /Bcosθ
(1)
where D is the average grain size, λ is the wave length of the incident X ray (1.5418 Å in the current research) and B is the width of the diffraction peak at half maximum for the diffraction angle 2θ, θ is the diffraction angle. Particles’ morphology was observed with a scanning electron microscope (SEM, Philips FEI Siron). Specimens for electric resistivity measurement were fabricated by pressing FeSi powder into a mold with a pressure of 750 MPa and holding for 90 s. Fabricated specimens were 10 mm in diameter and around 2 mm in thickness. Electric resistivity was measured by a four-probe technique at room temperature on a source meter (Keithley 2400). For each condition, 3 specimens were used to measure the electric resistivity and the mean values were used in the study. Magnetic properties were measured with a Lake Shore 7404 vibrating sample magnetometer (VSM) at room temperature. Electromagnetic properties of specimens containing 75 wt% of FeSi particles as filler and 25 wt% of paraffin as matrix were measured on a vector network analyzer (VNA, Agilent N5230A) at 2–18 GHz band. The VNA specimens were coaxial toroidal with outer diameter 7 mm, inner diameter of 3 mm and thickness of 3–3.5 mm. Microwave absorption performance can be evaluated by the transmission line theory [4,16], and the related formulas are as follows:
(
Zin = Z 0 μ r /εr
)1/2
tanh ⎡⎣j (2πfd/c ) (μ r εr )1/2⎤⎦
(2)
RL = 20 log (Zin − Z 0 )/(Zin + Z 0 )
153
(3)
where μr and εr are the effective permeability and permittivity of composite specimen, respectively, f is the frequency of electromagnetic wave, d is the thickness of absorber layer, c is the velocity of light, Z0 is the impedance of free space, and Zin is the input impedance of absorber layer.
3. Results and discussion 3.1. Phase structure and morphology Fig. 1 shows the x-ray diffraction patterns of FeSi alloy particles ball-milled for different time and milled-annealed. Three diffraction peaks are indexed to the (110), (200) and (211) planes of cubic α-Fe (Si) solid-solution, as is shown in the figure. No other diffraction peak can be identified, which indicates that neither phase transformation nor chemical reaction occurs during the ball-milling. Diffraction peaks become wider but the intensity decreases as the ball-milling time increases, as shown in the spectra, suggesting that the milling induces refined grains (more details about grain size is displayed in Table 1) and increased defects density. Similar phenomenon was previously observed in Refs. [17,18]. As shown in Table 1, the mean grain size decreases sharply from 39.4 nm to 17.3 nm in the first 10 h and remains at the level as the ballmilling further proceeds. Since the diffraction peak width increases persistently all through the ball-milling process despite the suspended grain refining, the broadening of diffraction peaks at the latter stage is believed attributed to the increase in defects density. On the other hand, diffraction peaks are found shift to higher angles as the ball-milling time increases, which suggests a decrease in the lattice size. The variation is believed results from the increase in the Si atoms’ dissolving into the Fe matrix. Since the radius of Si atom (0.118 nm) is smaller than that of Fe atom (0.124 nm), it’s replace of Fe induces a shrinkage distortion. This shrinkage enhances when increased replace occurs as the ballmilling progresses. Diffraction peaks of H500r20h are narrower and higher, comparing to those observed in XRD pattern of ON500r20h. Also, a peaks’ shift back to lower angle is observed, as shown in the figure. This reversion in the XRD pattern reveals that the annealing can help to reduce the defect density, release the lattice distortion and then perfect the crystal structure of particles. The morphology of FeSi alloy particles of different condition is presented in Fig. 2. Raw particles are spherical with diameter ranging from 5 to 30 μm. Once milled, particles are flattened into flakes with increased diameter and decreased thickness. When the ball-milling time approaches to 10–12 h, flaky particles with the highest aspect ratio are fabricated. Max diameter of around 70 μm and the mean thickness around 1 μm is obtained, as seen in Fig. 2 (c). The thin thickness is helpful to suppress the eddy current effects since it is near or even below the skin depth (about 1 μm reported by Ref [19]).When the ball-milling further proceeds, flaky particles are struck and sheared to fine pieces with reduced diameters of 4–20 μm and thickness of a few microns. Fig. 2(e) shows that, some small pieces attach to each other and the surface of particles is quite rough. Through the process, the aspect ratio is found increases firstly and then decreases with the highest aspect ratio achieved through 12 h milling. Particles’ shape and size changes little during the annealing, according to the comparison between the morphology in Fig. 2(e) and (f). 3.2. Electrical resistivity The dc electrical resistivity of FeSi alloy particles are listed in
154
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Fig. 1. XRD patterns of FeSi alloy particles milled for 0 h, 5 h, 10 h, 20 h and mill-annealed, (a) 2 theta range from 10º to 90º and (b) 2 theta range from 43º to 47º. Table 1 Grain size calculated by the (110) x-ray diffraction peak
Table 2 Electrical resistivity for FeSi alloy particles milled for various time or milled and then annealed.
Samples
Raw
ON500r5h
ON500r10h
ON500r20h
H500r20h
Grain size (nm)
39.4
20.4
17.3
16.0
51.8
Table 2. All the values of resistivity are about 3 to 4 orders of magnitude higher than that of bulk metal materials, due to the high contact resistance between particles. As we can see, the electrical resistivity increases as the ball-milling proceeds. After annealed, the resistivity of ON500r20h sample reduces sharply from 0.022 to 0.0039 Ω.cm. The increased surface atoms content together with the increased defects density that developed during the ball-milling induce an enhanced electron scattering which in turn contributes to the increase in the resistance. The recovery microstructure evolution during the annealing contributes to the decrease of the resistance.
Samples
ON500r10h
ON500r14h
ON500r20h
H500r20h
Electrical resistivity (Ω cm)
0.0081
0.016
0.022
0.0039
3.3. Static magnetic properties The magnetic hysteresis loops of FeSi alloy particles, as-milled and milled-annealed, are shown in Fig. 3 and the corresponding saturation magnetization (Ms), coercivity (Hc) and remanent magnetization (Mr) are listed in Table 3. As is displayed in Table 3, Ms of FeSi powder doesn’t change apparently until the ball-milling proceeds to 12 h and then reduces quickly from 202.1 to 138.0 emu/g after ball-milled for 20 h. The drop can be related to two factors, (i) the increased defects density [7], as confirmed by the broadening of XRD peak; (ii) the enhanced dissolution of Si into the Fe matrix [17], as confirmed by XRD peaks’ shifting to
Fig. 2. SEM micrographs of raw FeSi alloy particles (a) and particles ball-milled for 5 h (b), 10 h (c), 12 h (d), 20 h (e) and milled-annealed (f).
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Fig. 3. Magnetic hysteresis loops of FeSi alloy particles milled for different time or milled-annealed. Table 3 The saturation magnetization (Ms), the coercivity (Hc) and the remanent magnetization (Mr) of FeSi alloy particles milled for various time or milled and then annealed. Samples
Ms (emu/g)
Hc (Oe)
Mr (emu/g)
ON500r05h ON500r10h ON500r12h ON500r20h H500r20h
202.2 197.7 202.1 138.0 204.9
40.0 46.4 55.9 65.4 60.6
2.2 5.8 8.3 5.6 8.4
higher angle. The coercivity of the particles increases persistently as the ball-milling proceeds. Flaky particles with rough surface have higher shape anisotropy and high surface anisotropy, which is considered to be responsible for the increase in coercivity [20]. On the other hand, the decreased grain size as well as the increased defects density induces a higher resistance to the magnetization due to the pinning effect, which will also contribute to the increase of coercivity [11]. Mr as first increases from 2.2 emu/g to 8.3 emu/g at the first 12 h and then drops to 5.6 emu/g at 20 h, as listed in the table. The increase at the early stage is due to the increased coercivity and the latter decrease is attributed to the decrease in the saturation magnetization during the 12–20 h span. Ms recovers from 138.0 to 204.9 emu/g and Hc decreases slightly from 65.4 to 60.6 Oe after the as-milled particles were annealed, as shown in the Table 3. The reduced defects density, the decreased Si solving, the increase of grain size and the perfection of crystal structure that occurred during the annealing are believed to contribute to the recovery of the magnetic properties.
measured resistivity is of 10-1–10-2 Ω cm range, which agrees well with the calculation. Besides, multiple dispersion peaks can be observed in Fig. 4(a) and (b). The behavior is believed attributed to a conductivity gradient or difference in conductivity near the surface of the particle compared to its core [22], considering that ball-milling process might have changed the microstructure in the surface layer of the particles. On the other hand, both ε’ and ε” first increase and then decrease as the ball-milling proceeds, with an highest complex permittivity achieved as particles milled for 12 h are used as fillers. Interfacial polarization between the conductive metal and the insulating matrix primarily determines the permittivity of metalloaded composite, as reported previously [11,23–25]. Flaky particles possess higher aspect ratio and thus larger specific surface area comparing with spherical ones, which then contributes to an enhanced overall interfacial polarization. The increased electrical resistivity of the particles may contribute to the drop of permittivity observed when particles milled for more than 12 h as it may hinder the transfer of charges. Comparing the permittivity of ON500r20h to that of H50020h, the annealing is found cause a further drop in the permittivity. The annealing will enhance structure homogeneity insider particles such as the increase of grain size or decrease of defects density, which results in the reducing of dipoles [26]. Frequency dependence of the real part (μ’) and the imaginary part (μ”) of complex permeability for composite containing FeSi particles of different conditions are shown in Fig. 4(c) and (d), respectively. The μ’ increases apparently in the 2–6 GHz band while decreases in the higher frequency range (6–18 GHz) as the ball-milling proceeds during the first 12 h. A maximum μ’ of 5.4 is observed in specimen using powders ON500r12h as fillers. This variation change over when the ball-milling further proceeds. The variation of μ” with ball-milling time is quite similar to that of the μ’ and the maximum μ” (3.6, at 4.9 GHz) is also observed in ON500r12h. Besides, a peak in the curve of μ” becomes evident during the first 12 h ball-milling, indicating an enhanced resonance in the 2–10 GHz band. The natural resonance basically dominates the complex permeability within the microwave band. The natural resonance frequency fr for a spheroid particle can be represented as follows [27]:
f r2 =
(
)(
γ 2 Hk + 4πMs (D⊥ − De ) Hk + 4πMs (Dh − De ) 2π 2π
) (4)
where Hk and γ are the magneto crystalline anisotropy field and gyromagnetic ratio, is the perpendicular demagnetization factor, D⊥ and De are the parallel demagnetization factors along with hard and easy axis, respectively. For a ferromagnetic crystal with the easy magnetization axis (100), Hk can be expressed through the following equation:
3.4. Electromagnetic properties Frequency dependence of the real part (ε’) and the imaginary part (ε”) of complex permittivity of the specimens containing FeSi particles are showed in Fig. 4(a) and (b). As is shown, ε’ decreases while ε” increases as the frequency increases for milled particles, which indicates an enhanced dielectric relaxation in the 1–18 GHz band to compare with specimen containing raw FeSi particles. The dielectric relaxation frequency (frel) of metal-coated particles was found related to the metal’s electrical resistivity. Calculation based on Maxwell–Garnett model shows that when the metal’s electrical resistivity increases from 10-5 to 10-1–102 Ω cm the frel shifts from 1017 Hz to 109 –1010 Hz [21]. The frel in the current case is located at the 109 –1010 Hz band and the
155
Hk =
2K1 μ 0 Ms
(5)
where K1 and μ0 are the magneto crystalline anisotropy constant and static initial susceptibility. Through Eq. (5), Hk is mainly related to Ms since the other two parameters usually don’t change. In Eq. (4), the parts, 4πMs (D⊥ De) and 4πMs (Dh De), reflect the effects of particles shape and size on the natural resonance frequency. For the oblate spheroid particle, the demagnetization factors can be expressed as follows [27]:
D⊥ =
⎛ ⎜1 − − 1 ⎜⎝
a r2 a r2
a r2
⎞ a r2 − 1 ⎟ 1 arcsin ⎟ ar −1 ⎠
(6)
156
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Fig. 4. The frequency dependence of electromagnetic parameters of FeSi alloy particles of different conditions. (a) The real part of complex permittivity, (b) the imaginary part of complex permittivity, (c) the real part of complex permeability and (d) the imaginary part of complex permeability.
Dh = De =
1 − D⊥ 2
(7)
where ar is the aspect ratio of particles’ size to thickness. Combining Eqs. (7) and (2), fr for an oblate spheroid particle can be written as follows:
f r2 =
(
γ 2 Hk Hk + 2πMs (3D⊥ − 1) 2π 2π
) (8)
According to Eq. (8), the natural resonance frequency is mainly determined by the magneto crystalline anisotropy field and the shape anisotropy field which is represented by the part of 2πMs (3D⊥ − 1). On the other hand, the perpendicular demagnetization factor is calculated through Eq. (6). As seen in Fig. 5, the demagnetization in perpendicular axis increases when the aspect ratio increases. Therefore, when the spherical particles with low aspect ratio are flattened into flaky shape with higher aspect ratio, D⊥ increases, resulting in a higher fr and an enhanced natural resonance at the low-frequency end of the 2–18 GHz band. Specifically, most intensive resonance is observed as the particles milled for 12 h were used as fillers, as demonstrated in Fig. 4(c) and (d). When the ball-milling further proceeds, D⊥ decreases as the flaky particles fracture into small pieces, Hk increases as Ms decreases according to Eq. (4). All these variations lead to the resonance frequency’s shifting towards lower frequency and simultaneously a weakened natural resonance, resulting in a decreased permeability in low frequency region of 2–5 GHz band. The eddy current and the related skin effects also influence the complex permeability of conductive particles. The relationship between the skin depth δ and the particles conductivity σ can be
Fig. 5. Perpendicular demagnetization factors for oblate particles as a function of aspect ratio.
expressed as the equation [28]
δ=
1 πμi fσ
(9)
where f and μi are the frequency and the intrinsic permeability respectively. When spherical particles are gradually flatten into flaky particles during the first 12 h, the thin thickness together with the reduced conductivity suppress the negative influence
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
157
Fig. 6. The frequency dependence of the reflection loss for FeSi alloy particles, (a) raw commercial particles, (b) milled for 12 h, (c) milled for 20 h and (d) milled for 20 h and then annealed for another 20 h.
from the eddy current and the related skin effects effectively, which contributes partly to the increase of the complex permeability. After annealed, the permeability, especially the μ′ ′, decreases, which is partially related to the reinforcement of the eddy current effects. 3.5 EMA performance Fig. 6 shows the frequency dependency of reflection loss for EMA coatings containing FeSi alloy particles of different conditions as fillers and the related parameters are listed in Table 4. As is shown in Fig. 6, the location of RL peaks is influenced by particles’ shape. Compared with the coating (2 mm in thickness) containing raw FeSi particles, the RL peak of which locates at 16.8 GHz, coatings using flaky FeSi particles as fillers present high Table 4 The microwave absorption properties for FeSi alloy particles of different conditions Samples
Frequency of the lowest absorption peak (GHz)
Raw 16.8 ON500005h 2.8 ON500r10h o 2 ON500r12h o 2 ON500r20h o 2 H500r20h o2
Lowest The matching RL value (dB) thickness dm (mm)
Frequency of absorption peak at dm ¼ 1.5 mm (GHz)
Minimum RL value at dm ¼ 1.5 mm (dB)
55.88 21.22 9.67 10 19.27 30.44
– 11.1 2.9 2.4 3.9 5.8
– 10.45 7.8 8.84 11.05 13.05
2.01 4.89 2.22 1.88 3.03 3.86
EMA efficiency in lower frequency band (1–4 GHz). In addition, the location ( fm ) of the RL peak in 1.5 mm coatings is found shift first to the lower and then to higher frequency, as seen from Table 4, as the spherical particles get flattened and then broken into pieces. Specifically, RL of the coating using ON500r12h as fillers is lower than 5dB all through 2–4 GHz band, which is of technical significance. The locates of RL peaks is related to the electromagnetic properties of the coating at set thickness, according to the formula [29]
dm =
λm c = 4 4fm εr μ r
(10)
where dm is the matching thickness of the coating, it is equal to λ m /4 , λ m is the wavelength of microwave and c is the light speed. It’s noted that fm will decrease if the product of permittivity and permeability increases. Milled particles with higher aspect ratio, possessing higher electromagnetic parameters, endue lower fm than ones with lower aspect ratio. Additionally, microstructure of particles other than shape also influence microwave absorption performance. The coating containing ON500r20h as fillers possesses a RL peak of 19.27 dB at 2 GHz with the thickness of 3.03 mm. As comparison, the coating containing H500r20h presents a RL peak of 23.00 dB at 2.3 GHz with the thickness of 3.00 mm. Annealing induces a slight shift of fm to higher frequency when similar thickness is applied. Since the thickness and shape change slightly during the annealing, microstructure evolution including grain size and defects density is
158
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
believed affect microwave absorption performance. But the effect related to the microstructure is relatively weak to compare with that of shape.
4. Conclusions In this work, flaky FeSi alloy particles with different aspect ratio were produced via controlling the ball-milling process. A postmilling annealing was utilized to further improve the EM properties of the powder. It’s found that particles with higher aspect ratio possess higher permittivity and permeability. Besides, grain size, defects density and other changes inside particles are also found affect particles’ electromagnetic properties. Increased permittivity and permeability compels the frequency of absorption peak’s shift to lower frequency. Specifically, when using FeSi particles milled for 12 h as fillers, the coating of 1.5 mm present a RL below 5 dB all through the 2–4 GHz band, which is of technical significance for solving EMI/EMP problems in the band of wireless communication. The current study thus presents a feasible method to fabricate high performances EMA fillers in large quantity for application in the L–S band.
[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]
Acknowledgment This work was financially supported by the National Natural Science Foundation of China (Grant no. 51201048). J.T. Jiang thanks the Ph.D. Programs Foundation of Ministry of Education of China (Grant no. 20112302120021) Shanghai Aerospace Science and Technology Innovation Fund for the support.
References [1] K. Haga, S. Sugimoto, T. Kagotani, K. Inomata, Inomata, J. Akedo, Mater. Trans. 45 (2004) 2606–2609. [2] X. Tang, Q. Tian, B.Y. Zhao, K.A. Hu, Mater. Sci. Eng. A 135 (2007) 445–446.
[24] [25] [26] [27] [28] [29]
S.S. Kim, S.T. Kim, Y.C. Yoon, K.S. Lee, J. Appl. Phys. 97 (2005) 10F905. Y. Naito, K. Suetake, IEEE Trans. Microwave Theory Tech. 19 (1971) 65–72. J.L. Snoek, Nature 159 (1947) 90. W.F. Yang, L. Qiao, J.Q. Wei, Z.Q. Zhang, T. Wang, F.S. Li, J. Appl. Phys. 107268 (2010) 033913. R. Han, L. Qiao, T. Wang, F.S. Li, J. Alloys Compd. 509 (2011) 2734–2737. R. Han, X.H. Han, L. Qiao, T. Wang, F.S. Li, Mater. Chem. Phys. 128 (2011) 317–322. M.G. Han, D.F. Liang, K.N. Rozanov, L.J. Deng, IEEE Trans. Magn. 49 (2013) 982–985. M. Matsumoto, Y. Miyata, IEEE Trans. Magn. 33 (1997) 4459–4464. Y.X. Gong, L. Zhen, J.T. Jiang, C.Y. Xu, W.Z. Shao, J. Appl. Phys. 106 (2009) 064302. R.M. Walser, W. Kang, IEEE Trans. Magn. 34 (1998) 1144–1146. M.G. Han, W. Tang, W.B. Chen, H. Zhou, L.J. Deng, J. Appl. Phys. 107 (2010) 09A958. R.B. Yang, W.F. Liang, J. Appl. Phys. 113 (2013) 17A315. L.D. Liu, Y.P. Duan, J.B. Guo, L.Y. Chen, S.H. Liu, Physica B 406 (2011) 2261–2265. S.S. Kim, S.B. Jo, K.I. Gueon, K.K. Choi, J.M. Kim, IEEE Trans. Magn. 27 (1991) 5462–5466. Y.P. Duan, S.C. Gu, Z.L. Zhang, M. Wen, J. Alloys Compd. 542 (2012) 90–96. X. Wang, R.Z. Gong, P.G. Li, L.Y. Liu, W.M. Cheng, Mater. Sci. Eng. A 466 (2007) 178–182. G.X. Tong, J.H. Yuan, J. Ma, J.G. Guan, W.H. Wu, L.C. Li, R. Qiao, Mater. Chem. Phys. 129 (2011) 1189–1194. Q. Zeng, I. Baker, V. McCreary, Z.C. Yan, J. Magn. Magn. Mater. 318 (2007) 28–38. I.J. Youngs, N. Bowler, K.P. Lymer, S. Hussian, J. Phys. D Appl. Phys. 38 (2005) 188–201. N. Bowler, J. Phys. D: Appl. Phys. 37 (2004) 326–333. Y.X. Gong, L. Zhen, J.T. Jiang, C.Y. Xu, W.Z. Shao, J. Magn. Magn. Mater. 321 (2009) 3702–3705. N. Bowler, IEEE Trans. Dielectr. Electr. Insul. 13 (2006) 703–710. Y.B. Feng, T. Qiu, J. Magn. Magn. Mater. 324 (2012) 2528–2533. G.X. Tong, W.H. Wu, J.G. Guan, H.S. Qian, J.H. Yuan, W. Li, J. Alloys Compd. 509 (2011) 4320–4326. R.M. Walser, W. Win, P.M. Valanju, IEEE Trans. Magn. 34 (1998) 1390–1392. L.Z. Wu, J. Ding, H.B. Jiang, L.F. Chen, C.K. Ong, J. Magn. Magn. Mater. 285 (2005) 233–239. Y.P. Duan, Y.H. Zhang, T.M. Wang, S.C. Gu, X. Li, X.J. Lv, Mater. Sci. Eng. B-Adv. 185 (2014) 86–93.
Contents lists available at ScienceDirect
Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm
Microwave absorption properties of FeSi flaky particles prepared via a ball-milling process Chao Liu a, Yong Yuan b, Jian-tang Jiang a, Yuan-xun Gong c, Liang Zhen a,d,n a
School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China Precision Machinery Research Institute of Shanghai Space Flight Academy, Shanghai 201600, China c Aerospace Research Institute of Special Material and Processing Technology, Beijing 100074, China d MOE Key Laboratory of Micro-system and Micro-structures Manufacturing, Harbin Institute of Technology, Harbin 150080, China b
art ic l e i nf o
a b s t r a c t
Article history: Received 23 January 2015 Received in revised form 26 May 2015 Accepted 27 June 2015 Available online 8 July 2015
Flaky FeSi alloy particles with different aspect ratio were produced via ball-milling and a subsequent annealing. The microstructure and the morphology of the particles were examined by XRD and SEM. The dc resistivity, the static magnetization properties and electromagnetic properties were measured. Particles with high aspect ratio were found possess high permittivity and permeability. On the other hand, the variation of grain size and defects density was found influence the permittivity and permeability. High specific area was believed contribute to the intense dielectric loss and the high shape magnetic anisotropy lead to high permeability in the target band. Increased electromagnetic parameters compel the absorption peak’s shift to lower frequency. Coating using flaky FeSi particles milled for 12 h as fillers presented a reflection loss of 10 dB at 2 GHz and a matching thickness of 1.88 mm. The flaky FeSi alloy particles prepared through ball-milling and annealing can be promising candidates for EMA application at 1–4 GHz band. & 2015 Elsevier B.V. All rights reserved.
Keywords: FeSi particles Ball-milling Shape Microstructure Electromagnetic properties
1. Introduction With the fast development of wireless communication, the problem of electromagnetic interference (EI) and electromagnetic pollution (EP) has become more and more serious, especially in the 1–4 GHz band at which most communication systems work [1]. To absorb the harmful electromagnetic wave is an effective way to eliminate EI and EP problems, which basically rely on the application of electromagnetic wave absorbing (EMA) materials. Ferromagnetic metal/alloy powder of micro scale is the most promising candidate for EMA applications in the 1–4 GHz band since this series of materials possesses a unique potency of presenting a combination of high permeability and high permittivity [2]. To exert the potency of this series of materials is then a basic step to design high-performance EMA materials. High and commensurate electromagnetic (EM) properties (the complex permeability and the complex permittivity) ensure small matching thickness and good impedance matching [3,4], which is crucial for obtaining high EMA performance in 1–4 GHz band. A lot of researches have then been carried out to obtain high EM n Corresponding author at: School of Material Science and Engineering, Harbin Institute of Technology, Harbin 150001, China. Fax: þ 86 451 8641 3922. E-mail address: [email protected] (L. Zhen).
http://dx.doi.org/10.1016/j.jmmm.2015.06.085 0304-8853/& 2015 Elsevier B.V. All rights reserved.
properties. Among these efforts, particles with flaky shape and thin thickness have attracted intense attention, since they provide a possibility to obtain high permeability in 1–4 GHz band. The Snoek’s limit demonstrates [5,6] that it is essentially difficult to increases the permeability and the resonance frequency (fr) simultaneously, revealing the obstacle to achieve high permeability in GHz band. This limit, however, could be exceeded by introducing easy-plane anisotropy via using flaky particles [7–10]. Additionally, thin thickness is believed helpful to suppress the negative effects from the eddy current in metal or alloys particles [11, 12]. The eddy current induces reversed magnetization in particles and also reduces the effective volume due to the skin effect, which then leads to deteriorated permeability. Han et al. [13] flattened spherical iron particles into fine flakes and observed that particles with aspect ratio (AR) of 5 presented much higher permeability than that of spherical ones. A minimum reflection loss (RL) of 26 dB was observed at 3.5 GHz in a thin (2.0 mm) coating using flaky iron particles as fillers. Yang [14] prepared flaky Fe16Ni82Mo2 particles through a similar method and wherefrom achieved increased permeability and permittivity. The corresponding coating with a thickness of 2 mm presented a RL of 20.4 dB at 13.0 GHz. Liu et al. [15] prepared spherical FeSi alloy powders of various size by gas-atomization method and found that both the permeability and the permittivity increased as the particle size decreased. Among these studies, ball-milling was
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
found an efficient method to prepare flaky metal/alloy particles and improved EMA performances were wherefrom obtained. Most attention has been focused on the influence related to particles’ shape/size in the previous research, the evolution in microstructure characteristics including the grain size, the defects density and so on, however has not been investigated systematically. In this paper, flaky FeSi alloy particles were produced by ball-milling commercial FeSi alloy powders in large quantity to examine the feasibility. Post-milling annealing was thereafter performed to tailor the microstructure of the powder. The morphology and microstructure of the particles were examined and their influence on EM properties and the EMA performance were investigated.
2. Experimental details Commercially available gas-atomized FeSi alloy powder with mesh -400 was used as the raw material of the current study. Composition analysis indicated that the starting powder contain 6.31 wt% of Si, 0.07 wt% of Cr and allowance of Fe. The raw powder, certain content anhydrous ethanol and GCr15 steel balls were added into jars of 250 ml and then the ball-milling process was conducted in a planetary ball-mill at a rotation rate of 500 rpm for up to 20 h. Through the study, the ball–powder–ethanol ratio was set at 5:1:1. After the ball-milling completed, the product was collected, washed for three times with anhydrous ethanol and then dried for 8 h at 50 ºC in an oven. The product milled for 20 h was labeled ON500r20h for short and samples of other conditions were labeled in the same way. Some as-milled powder (ON500r20h) was transferred into annealing process. The power was heated in a furnace in hydrogen atmosphere at 800 ºC, hold for 2 h and then let cooled to room temperature naturally. The annealed powder was labeled as H500r20h. Phase structure of the particles, raw, milled and mill-annealed, was examined on a PANalytical X’Pert PRO diffractometer (XRD) using Cu Kα radiation. The grain size of the particles was calculated according to Scherer formula
D = Kλ /Bcosθ
(1)
where D is the average grain size, λ is the wave length of the incident X ray (1.5418 Å in the current research) and B is the width of the diffraction peak at half maximum for the diffraction angle 2θ, θ is the diffraction angle. Particles’ morphology was observed with a scanning electron microscope (SEM, Philips FEI Siron). Specimens for electric resistivity measurement were fabricated by pressing FeSi powder into a mold with a pressure of 750 MPa and holding for 90 s. Fabricated specimens were 10 mm in diameter and around 2 mm in thickness. Electric resistivity was measured by a four-probe technique at room temperature on a source meter (Keithley 2400). For each condition, 3 specimens were used to measure the electric resistivity and the mean values were used in the study. Magnetic properties were measured with a Lake Shore 7404 vibrating sample magnetometer (VSM) at room temperature. Electromagnetic properties of specimens containing 75 wt% of FeSi particles as filler and 25 wt% of paraffin as matrix were measured on a vector network analyzer (VNA, Agilent N5230A) at 2–18 GHz band. The VNA specimens were coaxial toroidal with outer diameter 7 mm, inner diameter of 3 mm and thickness of 3–3.5 mm. Microwave absorption performance can be evaluated by the transmission line theory [4,16], and the related formulas are as follows:
(
Zin = Z 0 μ r /εr
)1/2
tanh ⎡⎣j (2πfd/c ) (μ r εr )1/2⎤⎦
(2)
RL = 20 log (Zin − Z 0 )/(Zin + Z 0 )
153
(3)
where μr and εr are the effective permeability and permittivity of composite specimen, respectively, f is the frequency of electromagnetic wave, d is the thickness of absorber layer, c is the velocity of light, Z0 is the impedance of free space, and Zin is the input impedance of absorber layer.
3. Results and discussion 3.1. Phase structure and morphology Fig. 1 shows the x-ray diffraction patterns of FeSi alloy particles ball-milled for different time and milled-annealed. Three diffraction peaks are indexed to the (110), (200) and (211) planes of cubic α-Fe (Si) solid-solution, as is shown in the figure. No other diffraction peak can be identified, which indicates that neither phase transformation nor chemical reaction occurs during the ball-milling. Diffraction peaks become wider but the intensity decreases as the ball-milling time increases, as shown in the spectra, suggesting that the milling induces refined grains (more details about grain size is displayed in Table 1) and increased defects density. Similar phenomenon was previously observed in Refs. [17,18]. As shown in Table 1, the mean grain size decreases sharply from 39.4 nm to 17.3 nm in the first 10 h and remains at the level as the ballmilling further proceeds. Since the diffraction peak width increases persistently all through the ball-milling process despite the suspended grain refining, the broadening of diffraction peaks at the latter stage is believed attributed to the increase in defects density. On the other hand, diffraction peaks are found shift to higher angles as the ball-milling time increases, which suggests a decrease in the lattice size. The variation is believed results from the increase in the Si atoms’ dissolving into the Fe matrix. Since the radius of Si atom (0.118 nm) is smaller than that of Fe atom (0.124 nm), it’s replace of Fe induces a shrinkage distortion. This shrinkage enhances when increased replace occurs as the ballmilling progresses. Diffraction peaks of H500r20h are narrower and higher, comparing to those observed in XRD pattern of ON500r20h. Also, a peaks’ shift back to lower angle is observed, as shown in the figure. This reversion in the XRD pattern reveals that the annealing can help to reduce the defect density, release the lattice distortion and then perfect the crystal structure of particles. The morphology of FeSi alloy particles of different condition is presented in Fig. 2. Raw particles are spherical with diameter ranging from 5 to 30 μm. Once milled, particles are flattened into flakes with increased diameter and decreased thickness. When the ball-milling time approaches to 10–12 h, flaky particles with the highest aspect ratio are fabricated. Max diameter of around 70 μm and the mean thickness around 1 μm is obtained, as seen in Fig. 2 (c). The thin thickness is helpful to suppress the eddy current effects since it is near or even below the skin depth (about 1 μm reported by Ref [19]).When the ball-milling further proceeds, flaky particles are struck and sheared to fine pieces with reduced diameters of 4–20 μm and thickness of a few microns. Fig. 2(e) shows that, some small pieces attach to each other and the surface of particles is quite rough. Through the process, the aspect ratio is found increases firstly and then decreases with the highest aspect ratio achieved through 12 h milling. Particles’ shape and size changes little during the annealing, according to the comparison between the morphology in Fig. 2(e) and (f). 3.2. Electrical resistivity The dc electrical resistivity of FeSi alloy particles are listed in
154
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Fig. 1. XRD patterns of FeSi alloy particles milled for 0 h, 5 h, 10 h, 20 h and mill-annealed, (a) 2 theta range from 10º to 90º and (b) 2 theta range from 43º to 47º. Table 1 Grain size calculated by the (110) x-ray diffraction peak
Table 2 Electrical resistivity for FeSi alloy particles milled for various time or milled and then annealed.
Samples
Raw
ON500r5h
ON500r10h
ON500r20h
H500r20h
Grain size (nm)
39.4
20.4
17.3
16.0
51.8
Table 2. All the values of resistivity are about 3 to 4 orders of magnitude higher than that of bulk metal materials, due to the high contact resistance between particles. As we can see, the electrical resistivity increases as the ball-milling proceeds. After annealed, the resistivity of ON500r20h sample reduces sharply from 0.022 to 0.0039 Ω.cm. The increased surface atoms content together with the increased defects density that developed during the ball-milling induce an enhanced electron scattering which in turn contributes to the increase in the resistance. The recovery microstructure evolution during the annealing contributes to the decrease of the resistance.
Samples
ON500r10h
ON500r14h
ON500r20h
H500r20h
Electrical resistivity (Ω cm)
0.0081
0.016
0.022
0.0039
3.3. Static magnetic properties The magnetic hysteresis loops of FeSi alloy particles, as-milled and milled-annealed, are shown in Fig. 3 and the corresponding saturation magnetization (Ms), coercivity (Hc) and remanent magnetization (Mr) are listed in Table 3. As is displayed in Table 3, Ms of FeSi powder doesn’t change apparently until the ball-milling proceeds to 12 h and then reduces quickly from 202.1 to 138.0 emu/g after ball-milled for 20 h. The drop can be related to two factors, (i) the increased defects density [7], as confirmed by the broadening of XRD peak; (ii) the enhanced dissolution of Si into the Fe matrix [17], as confirmed by XRD peaks’ shifting to
Fig. 2. SEM micrographs of raw FeSi alloy particles (a) and particles ball-milled for 5 h (b), 10 h (c), 12 h (d), 20 h (e) and milled-annealed (f).
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Fig. 3. Magnetic hysteresis loops of FeSi alloy particles milled for different time or milled-annealed. Table 3 The saturation magnetization (Ms), the coercivity (Hc) and the remanent magnetization (Mr) of FeSi alloy particles milled for various time or milled and then annealed. Samples
Ms (emu/g)
Hc (Oe)
Mr (emu/g)
ON500r05h ON500r10h ON500r12h ON500r20h H500r20h
202.2 197.7 202.1 138.0 204.9
40.0 46.4 55.9 65.4 60.6
2.2 5.8 8.3 5.6 8.4
higher angle. The coercivity of the particles increases persistently as the ball-milling proceeds. Flaky particles with rough surface have higher shape anisotropy and high surface anisotropy, which is considered to be responsible for the increase in coercivity [20]. On the other hand, the decreased grain size as well as the increased defects density induces a higher resistance to the magnetization due to the pinning effect, which will also contribute to the increase of coercivity [11]. Mr as first increases from 2.2 emu/g to 8.3 emu/g at the first 12 h and then drops to 5.6 emu/g at 20 h, as listed in the table. The increase at the early stage is due to the increased coercivity and the latter decrease is attributed to the decrease in the saturation magnetization during the 12–20 h span. Ms recovers from 138.0 to 204.9 emu/g and Hc decreases slightly from 65.4 to 60.6 Oe after the as-milled particles were annealed, as shown in the Table 3. The reduced defects density, the decreased Si solving, the increase of grain size and the perfection of crystal structure that occurred during the annealing are believed to contribute to the recovery of the magnetic properties.
measured resistivity is of 10-1–10-2 Ω cm range, which agrees well with the calculation. Besides, multiple dispersion peaks can be observed in Fig. 4(a) and (b). The behavior is believed attributed to a conductivity gradient or difference in conductivity near the surface of the particle compared to its core [22], considering that ball-milling process might have changed the microstructure in the surface layer of the particles. On the other hand, both ε’ and ε” first increase and then decrease as the ball-milling proceeds, with an highest complex permittivity achieved as particles milled for 12 h are used as fillers. Interfacial polarization between the conductive metal and the insulating matrix primarily determines the permittivity of metalloaded composite, as reported previously [11,23–25]. Flaky particles possess higher aspect ratio and thus larger specific surface area comparing with spherical ones, which then contributes to an enhanced overall interfacial polarization. The increased electrical resistivity of the particles may contribute to the drop of permittivity observed when particles milled for more than 12 h as it may hinder the transfer of charges. Comparing the permittivity of ON500r20h to that of H50020h, the annealing is found cause a further drop in the permittivity. The annealing will enhance structure homogeneity insider particles such as the increase of grain size or decrease of defects density, which results in the reducing of dipoles [26]. Frequency dependence of the real part (μ’) and the imaginary part (μ”) of complex permeability for composite containing FeSi particles of different conditions are shown in Fig. 4(c) and (d), respectively. The μ’ increases apparently in the 2–6 GHz band while decreases in the higher frequency range (6–18 GHz) as the ball-milling proceeds during the first 12 h. A maximum μ’ of 5.4 is observed in specimen using powders ON500r12h as fillers. This variation change over when the ball-milling further proceeds. The variation of μ” with ball-milling time is quite similar to that of the μ’ and the maximum μ” (3.6, at 4.9 GHz) is also observed in ON500r12h. Besides, a peak in the curve of μ” becomes evident during the first 12 h ball-milling, indicating an enhanced resonance in the 2–10 GHz band. The natural resonance basically dominates the complex permeability within the microwave band. The natural resonance frequency fr for a spheroid particle can be represented as follows [27]:
f r2 =
(
)(
γ 2 Hk + 4πMs (D⊥ − De ) Hk + 4πMs (Dh − De ) 2π 2π
) (4)
where Hk and γ are the magneto crystalline anisotropy field and gyromagnetic ratio, is the perpendicular demagnetization factor, D⊥ and De are the parallel demagnetization factors along with hard and easy axis, respectively. For a ferromagnetic crystal with the easy magnetization axis (100), Hk can be expressed through the following equation:
3.4. Electromagnetic properties Frequency dependence of the real part (ε’) and the imaginary part (ε”) of complex permittivity of the specimens containing FeSi particles are showed in Fig. 4(a) and (b). As is shown, ε’ decreases while ε” increases as the frequency increases for milled particles, which indicates an enhanced dielectric relaxation in the 1–18 GHz band to compare with specimen containing raw FeSi particles. The dielectric relaxation frequency (frel) of metal-coated particles was found related to the metal’s electrical resistivity. Calculation based on Maxwell–Garnett model shows that when the metal’s electrical resistivity increases from 10-5 to 10-1–102 Ω cm the frel shifts from 1017 Hz to 109 –1010 Hz [21]. The frel in the current case is located at the 109 –1010 Hz band and the
155
Hk =
2K1 μ 0 Ms
(5)
where K1 and μ0 are the magneto crystalline anisotropy constant and static initial susceptibility. Through Eq. (5), Hk is mainly related to Ms since the other two parameters usually don’t change. In Eq. (4), the parts, 4πMs (D⊥ De) and 4πMs (Dh De), reflect the effects of particles shape and size on the natural resonance frequency. For the oblate spheroid particle, the demagnetization factors can be expressed as follows [27]:
D⊥ =
⎛ ⎜1 − − 1 ⎜⎝
a r2 a r2
a r2
⎞ a r2 − 1 ⎟ 1 arcsin ⎟ ar −1 ⎠
(6)
156
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
Fig. 4. The frequency dependence of electromagnetic parameters of FeSi alloy particles of different conditions. (a) The real part of complex permittivity, (b) the imaginary part of complex permittivity, (c) the real part of complex permeability and (d) the imaginary part of complex permeability.
Dh = De =
1 − D⊥ 2
(7)
where ar is the aspect ratio of particles’ size to thickness. Combining Eqs. (7) and (2), fr for an oblate spheroid particle can be written as follows:
f r2 =
(
γ 2 Hk Hk + 2πMs (3D⊥ − 1) 2π 2π
) (8)
According to Eq. (8), the natural resonance frequency is mainly determined by the magneto crystalline anisotropy field and the shape anisotropy field which is represented by the part of 2πMs (3D⊥ − 1). On the other hand, the perpendicular demagnetization factor is calculated through Eq. (6). As seen in Fig. 5, the demagnetization in perpendicular axis increases when the aspect ratio increases. Therefore, when the spherical particles with low aspect ratio are flattened into flaky shape with higher aspect ratio, D⊥ increases, resulting in a higher fr and an enhanced natural resonance at the low-frequency end of the 2–18 GHz band. Specifically, most intensive resonance is observed as the particles milled for 12 h were used as fillers, as demonstrated in Fig. 4(c) and (d). When the ball-milling further proceeds, D⊥ decreases as the flaky particles fracture into small pieces, Hk increases as Ms decreases according to Eq. (4). All these variations lead to the resonance frequency’s shifting towards lower frequency and simultaneously a weakened natural resonance, resulting in a decreased permeability in low frequency region of 2–5 GHz band. The eddy current and the related skin effects also influence the complex permeability of conductive particles. The relationship between the skin depth δ and the particles conductivity σ can be
Fig. 5. Perpendicular demagnetization factors for oblate particles as a function of aspect ratio.
expressed as the equation [28]
δ=
1 πμi fσ
(9)
where f and μi are the frequency and the intrinsic permeability respectively. When spherical particles are gradually flatten into flaky particles during the first 12 h, the thin thickness together with the reduced conductivity suppress the negative influence
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
157
Fig. 6. The frequency dependence of the reflection loss for FeSi alloy particles, (a) raw commercial particles, (b) milled for 12 h, (c) milled for 20 h and (d) milled for 20 h and then annealed for another 20 h.
from the eddy current and the related skin effects effectively, which contributes partly to the increase of the complex permeability. After annealed, the permeability, especially the μ′ ′, decreases, which is partially related to the reinforcement of the eddy current effects. 3.5 EMA performance Fig. 6 shows the frequency dependency of reflection loss for EMA coatings containing FeSi alloy particles of different conditions as fillers and the related parameters are listed in Table 4. As is shown in Fig. 6, the location of RL peaks is influenced by particles’ shape. Compared with the coating (2 mm in thickness) containing raw FeSi particles, the RL peak of which locates at 16.8 GHz, coatings using flaky FeSi particles as fillers present high Table 4 The microwave absorption properties for FeSi alloy particles of different conditions Samples
Frequency of the lowest absorption peak (GHz)
Raw 16.8 ON500005h 2.8 ON500r10h o 2 ON500r12h o 2 ON500r20h o 2 H500r20h o2
Lowest The matching RL value (dB) thickness dm (mm)
Frequency of absorption peak at dm ¼ 1.5 mm (GHz)
Minimum RL value at dm ¼ 1.5 mm (dB)
55.88 21.22 9.67 10 19.27 30.44
– 11.1 2.9 2.4 3.9 5.8
– 10.45 7.8 8.84 11.05 13.05
2.01 4.89 2.22 1.88 3.03 3.86
EMA efficiency in lower frequency band (1–4 GHz). In addition, the location ( fm ) of the RL peak in 1.5 mm coatings is found shift first to the lower and then to higher frequency, as seen from Table 4, as the spherical particles get flattened and then broken into pieces. Specifically, RL of the coating using ON500r12h as fillers is lower than 5dB all through 2–4 GHz band, which is of technical significance. The locates of RL peaks is related to the electromagnetic properties of the coating at set thickness, according to the formula [29]
dm =
λm c = 4 4fm εr μ r
(10)
where dm is the matching thickness of the coating, it is equal to λ m /4 , λ m is the wavelength of microwave and c is the light speed. It’s noted that fm will decrease if the product of permittivity and permeability increases. Milled particles with higher aspect ratio, possessing higher electromagnetic parameters, endue lower fm than ones with lower aspect ratio. Additionally, microstructure of particles other than shape also influence microwave absorption performance. The coating containing ON500r20h as fillers possesses a RL peak of 19.27 dB at 2 GHz with the thickness of 3.03 mm. As comparison, the coating containing H500r20h presents a RL peak of 23.00 dB at 2.3 GHz with the thickness of 3.00 mm. Annealing induces a slight shift of fm to higher frequency when similar thickness is applied. Since the thickness and shape change slightly during the annealing, microstructure evolution including grain size and defects density is
158
C. Liu et al. / Journal of Magnetism and Magnetic Materials 395 (2015) 152–158
believed affect microwave absorption performance. But the effect related to the microstructure is relatively weak to compare with that of shape.
4. Conclusions In this work, flaky FeSi alloy particles with different aspect ratio were produced via controlling the ball-milling process. A postmilling annealing was utilized to further improve the EM properties of the powder. It’s found that particles with higher aspect ratio possess higher permittivity and permeability. Besides, grain size, defects density and other changes inside particles are also found affect particles’ electromagnetic properties. Increased permittivity and permeability compels the frequency of absorption peak’s shift to lower frequency. Specifically, when using FeSi particles milled for 12 h as fillers, the coating of 1.5 mm present a RL below 5 dB all through the 2–4 GHz band, which is of technical significance for solving EMI/EMP problems in the band of wireless communication. The current study thus presents a feasible method to fabricate high performances EMA fillers in large quantity for application in the L–S band.
[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]
Acknowledgment This work was financially supported by the National Natural Science Foundation of China (Grant no. 51201048). J.T. Jiang thanks the Ph.D. Programs Foundation of Ministry of Education of China (Grant no. 20112302120021) Shanghai Aerospace Science and Technology Innovation Fund for the support.
References [1] K. Haga, S. Sugimoto, T. Kagotani, K. Inomata, Inomata, J. Akedo, Mater. Trans. 45 (2004) 2606–2609. [2] X. Tang, Q. Tian, B.Y. Zhao, K.A. Hu, Mater. Sci. Eng. A 135 (2007) 445–446.
[24] [25] [26] [27] [28] [29]
S.S. Kim, S.T. Kim, Y.C. Yoon, K.S. Lee, J. Appl. Phys. 97 (2005) 10F905. Y. Naito, K. Suetake, IEEE Trans. Microwave Theory Tech. 19 (1971) 65–72. J.L. Snoek, Nature 159 (1947) 90. W.F. Yang, L. Qiao, J.Q. Wei, Z.Q. Zhang, T. Wang, F.S. Li, J. Appl. Phys. 107268 (2010) 033913. R. Han, L. Qiao, T. Wang, F.S. Li, J. Alloys Compd. 509 (2011) 2734–2737. R. Han, X.H. Han, L. Qiao, T. Wang, F.S. Li, Mater. Chem. Phys. 128 (2011) 317–322. M.G. Han, D.F. Liang, K.N. Rozanov, L.J. Deng, IEEE Trans. Magn. 49 (2013) 982–985. M. Matsumoto, Y. Miyata, IEEE Trans. Magn. 33 (1997) 4459–4464. Y.X. Gong, L. Zhen, J.T. Jiang, C.Y. Xu, W.Z. Shao, J. Appl. Phys. 106 (2009) 064302. R.M. Walser, W. Kang, IEEE Trans. Magn. 34 (1998) 1144–1146. M.G. Han, W. Tang, W.B. Chen, H. Zhou, L.J. Deng, J. Appl. Phys. 107 (2010) 09A958. R.B. Yang, W.F. Liang, J. Appl. Phys. 113 (2013) 17A315. L.D. Liu, Y.P. Duan, J.B. Guo, L.Y. Chen, S.H. Liu, Physica B 406 (2011) 2261–2265. S.S. Kim, S.B. Jo, K.I. Gueon, K.K. Choi, J.M. Kim, IEEE Trans. Magn. 27 (1991) 5462–5466. Y.P. Duan, S.C. Gu, Z.L. Zhang, M. Wen, J. Alloys Compd. 542 (2012) 90–96. X. Wang, R.Z. Gong, P.G. Li, L.Y. Liu, W.M. Cheng, Mater. Sci. Eng. A 466 (2007) 178–182. G.X. Tong, J.H. Yuan, J. Ma, J.G. Guan, W.H. Wu, L.C. Li, R. Qiao, Mater. Chem. Phys. 129 (2011) 1189–1194. Q. Zeng, I. Baker, V. McCreary, Z.C. Yan, J. Magn. Magn. Mater. 318 (2007) 28–38. I.J. Youngs, N. Bowler, K.P. Lymer, S. Hussian, J. Phys. D Appl. Phys. 38 (2005) 188–201. N. Bowler, J. Phys. D: Appl. Phys. 37 (2004) 326–333. Y.X. Gong, L. Zhen, J.T. Jiang, C.Y. Xu, W.Z. Shao, J. Magn. Magn. Mater. 321 (2009) 3702–3705. N. Bowler, IEEE Trans. Dielectr. Electr. Insul. 13 (2006) 703–710. Y.B. Feng, T. Qiu, J. Magn. Magn. Mater. 324 (2012) 2528–2533. G.X. Tong, W.H. Wu, J.G. Guan, H.S. Qian, J.H. Yuan, W. Li, J. Alloys Compd. 509 (2011) 4320–4326. R.M. Walser, W. Win, P.M. Valanju, IEEE Trans. Magn. 34 (1998) 1390–1392. L.Z. Wu, J. Ding, H.B. Jiang, L.F. Chen, C.K. Ong, J. Magn. Magn. Mater. 285 (2005) 233–239. Y.P. Duan, Y.H. Zhang, T.M. Wang, S.C. Gu, X. Li, X.J. Lv, Mater. Sci. Eng. B-Adv. 185 (2014) 86–93.