Phase formation process, microstructure and magnetic properties of Y-type hexagonal ferrite prepared by citrate sol–gel auto-combustion method

Materials Chemistry and Physics 98 (2006) 66–70

Phase formation process, microstructure and magnetic properties of Y-type hexagonal ferrite prepared by citrate sol–gel auto-combustion method Yang Bai, Ji Zhou ∗ , Zhilun Gui, Longtu Li State Key Lab of New Ceramics and Fine Processing, Department of Materials Science and Engineering, Tsinghua University, Beijing 100084, PR China Received 21 March 2005; received in revised form 19 July 2005; accepted 29 August 2005

Abstract Y-type hexagonal ferrites were prepared by citrate sol–gel auto-combustion method (CSAM). The phase formation process, microstructure and magnetic properties were investigated and compared with those of the samples made by conventional solid-state reaction method (SSRM). The selected material system has stoichiometric composition of Ba2 Zn1.2−x Cox Cu0.8 Fe12 O22 , where x was varied from 0 to 1.2. Phase formation was characterized by powder X-ray diffraction (XRD). The microstructure was observed via scanning electron microscopy (SEM). Frequency spectra of complex permeability were measured via an impedance analyzer. Results reveal auto-combusted powders are about 22 nm and very active. Phase formation temperature and sintering temperature are both 900 ◦ C, much lower than those in SSRM. The sintered samples have fine-grained microstructure. The samples exhibit better magnetic properties in hyper-frequency than those made by SSRM. As zinc content increases, the permeability µ is enhanced and the cut-off frequency drops. The magnetic properties are also correlated with the sintering temperature and microstructure. Permeability µ increases with the rise of sintering temperature. © 2005 Elsevier B.V. All rights reserved. PACS: 75.50.Gg Keywords: Y-type hexagonal ferrite; Citrate sol–gel auto-combustion method; Phase formation; Low temperature sintering; Magnetic properties

1. Introduction Due to the rapid development of information and communication technology, integration and miniaturization are the general trend for electronic equipments. As more components are integrated in smaller space, electromagnetic compatibility and anti-electromagnetic interference (EMI) become one of most challenging problems in electronic products. The importance of special surface mount components to resist EMI in integrated circuit is driven up. That promotes a great demand of the chip soft-magnetic components (including multi-layer chip inductors (MLCI) and chip electromagnetic interference filters) in hyperfrequency applications. The conventional materials for MLCI, such as NiZnCu spinel ferrites, cannot be used in the hyperfrequency range due to the limitation of their cut-off frequency [1]. Y-type hexagonal ferrites have planar magnetic anisotropy. Their cut-off frequency is about an order of magnitude higher

∗

Corresponding author. Tel.: +86 10 62772975; fax: +86 10 62772975. E-mail address: [email protected] (J. Zhou).

0254-0584/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2005.08.067

than that of spinel ferrites [2]. Y-type hexagonal ferrite exhibits excellent magnetic properties in hyper-frequency [3,4]. It is anticipated that the Y-type hexagonal ferrite will meet the need of soft magnetic materials for chip components in hyperfrequency. MLCI adopts low temperature cofired ceramics (LTCC) technology, which needs the media materials have proper sintering temperature to be cofired with electrode materials. Considering the price, conductivity and the oxidation resistance in high temperature synthesis, Ag is the best choice as electrode material for MLCI. Because the melting point of Ag is 961 ◦ C, the soft magnetic materials should be sintered under 900 ◦ C [5]. The sintering temperature of Y-type hexaferrites is over 1000 ◦ C in conventional method. The usual way to lower the sintering temperature is the addition of some compounds with low melting point, while magnetic properties drop at the same time. Citrate sol–gel auto-combustion (CSAM) is a novel method to fabricate highly active nano-sized powders [6]. Hence, materials can be sintered at relatively lower temperature than that in the solidstate reaction method (SSRM) [7]. Our experimental results show that Y-type hexagonal ferrite can be sintered under 900 ◦ C

Y. Bai et al. / Materials Chemistry and Physics 98 (2006) 66–70

without any compound addition using CSAM and co-fired with Ag electrode directly.

67

3. Results and discussion 3.1. Phase formation process and particle size

2. Experimental procedure Y-type hexagonal ferrites were prepared by CSAM. The experimental material system has stoichimetric composition of Ba2 Zn1.2−x Cox Cu0.8 Fe12 O22 , in which x was varied from 0 to 1.2. Unlike conventional SSRM, analytic reagent grade nitrates, Fe(NO3 )3 ·9H2 O, Co(NO3 )2 ·6H2 O, Zn(NO3 )2 ·6H2 O, Cu(NO3 )2 ·6H2 O and (CH3 COO)2 Ba·H2 O, were selected as raw materials. Citrate was slowly added to the mixed solution of nitrates as a chelator. Concentrated ammonia solution was dropped into the solution until the solution was neutral or slightly alkaline (pH 7–8). Then the solution became homogeneous stable sol. The gel was dried in an oven at 120 ◦ C. With the aid of a little amount of ethanol, the obtained dried gel was ignited in an open container. The dry gel was burnt in a self-propagating combustion manner. During auto-combustion, the burning gel expanded rapidly in volume and grew as a dendritic structure. Loose powders were finally formed. The auto-combusted powders were annealed at 900 ◦ C in the air, before ball milled in a ball mill for 24 h using stainless-steel balls and alcohol as media. The resulting powders were pressed in a stainless-steel die under a pressure of 7 MPa with 5 wt.% polyvinyl alcohol as binder-lubricant. The pressed toroidal samples (20 mm outside diameter, 10 mm inside diameter, about 3 mm thickness) were sintered in the temperature range of 850–1000 ◦ C for 4 h in air and then cooled in the furnace. Powder X-ray diffraction (XRD) was applied to characterize the phase composition of the powder samples treated under different conditions using a Rigaku X-ray diffractometer equipped with Cu K␣ radiation (λ = 0.15405 nm). The microstructure of the fracture surface of the sintered specimens was observed via a scanning electron microscopy (SEM). Frequency dependence of permeability was measured using a Hewlett Packard HP4291B impedance analyzer from 1 MHz to 1 GHz. The samples prepared by conventional method were also measured for comparison.

A series of XRD patterns (Fig. 1(a)) show the phase formation process of the samples prepared by CSAM, which differs from that of samples by SSRM (Fig. 1(b)). After auto-combustion, the main phases of resulting powders are simple compounds, such as spinel ferrites and BaCO3 . That indicates that Y-type phase cannot be formed directly after auto-combustion due to its complex crystal structure. According to the compositions, spinel ferrites include ZnFe2 O4 , CuFe2 O4 , CoFe2 O4 and Fe3 O4 . With the rise of annealing temperature, BaCO3 reacts with Fe3 O4 to form BaFe2 O4 . After annealed at 750 ◦ C, BaCO3 disappears and BaFe2 O4 become one of the main phases in the samples. While the temperature more increases, BaFe2 O4 react with other spinel ferrites to form complex compound, Y-type hexaferrite. When the annealing temperature rises to 800 ◦ C, Y-type hexagonal ferrite begins to form. Some BaFe2 O4 and spinel ferrites do not react completely and remain in the samples. The results show that well-defined Y-type phase is formed at 900 ◦ C. Fig. 2 indicates that there is a pure Y-type hexagonal ferrite crystalline phase for each composition in the samples. CSAM lowers the phase formation temperature of Y-type ferroxplana markedly whereas Y-type phase was formed over 1000 ◦ C in SSRM. The lattice parameter of the samples with different compositions was calculated and listed in Table 1. Lattice parameters a and c all decline with the increasing Co content due to the difference of ion radii between Zn2+ and Co2+ .

Fig. 1. The XRD spectra of the samples prepared by (a) citrate sol–gel auto-combustion method and (b) conventional method annealed at different temperatures, standard Y-type phase spectrum is also shown for comparison (S: spinel ferrite; Y: Y-type hexagonal ferrite; B: BaFe2 O4 ; C: BaCO3 ; I: Fe2 O3 ; Z: ZnO).

68

Y. Bai et al. / Materials Chemistry and Physics 98 (2006) 66–70 Table 1 Dependence of lattice parameters a and c on the Co content x

a

c

0 0.3 0.6 0.9 1.2

5.875 5.872 5.870 5.865 5.865

43.689 43.601 43.557 43.427 43.506

sized powders with high activity so that the phase formation temperature and sintering temperature are much lower than those in SSRM. 3.2. Microstructure and densities

Fig. 2. XRD spectra of the samples with different compositions prepared by CSAM and annealed at 900 ◦ C.

The particle size of the powders was calculated according to Scherrer equation. The auto-combusted powders have average size of about 22 nm. After annealed, the granularity will augment with the rise of temperature and prolongation of annealing time. Highly active powders can be obtained by controlling the annealing process strictly. The particle size of the sample annealed at 800 ◦ C is about 72 nm, while that of the sample annealed at 900 ◦ C is about 88 nm. Large surface energy endows the nano-

The microstructure of the samples made by CSAM is compared with that of the sample using SSRM in Fig. 3. The grain morphology for each samples are all platelike. The average grain size of the samples via CSAM is about 2 ␮m, much smaller than that of the sample prepared by SSRM. CSAM lowers the sintering temperature dramatically compared with SSRM. The nano-sized particles are more active during the sintering process. High activity of particles speeds the sintering process and lowers the sintering temperature. The samples can be sintered well at 900 ◦ C, and have fine-grained microstructure. Grains are compactly stack and uniform in size. The sintering temperature is about 100 ◦ C lower than that in SSRM. According to the SEM results, the microstructure is not

Fig. 3. The SEM photos of the Ba2 Zn1.2−x Cox Cu0.8 Fe12 O22 samples prepared by CSAM and sintered at 900 ◦ C with the compositions of (a) x = 0.0, (b) x = 0.6, and (c) x = 1.2, compared with the photo of (d) the x = 0.6 sample prepared by SSRM and sintered at 1000 ◦ C.

Y. Bai et al. / Materials Chemistry and Physics 98 (2006) 66–70

69

Table 2 Density of the samples treated at different conditions (g cm−3 ) x

Sol–gel method Ts

0 0.3 0.6 0.9 1.2

= 900 ◦ C

5.1 5.1 5.0 5.1 5.1

Solid-state reaction method Ts

= 950 ◦ C

5.3 5.3 5.3 5.3 5.3

Ts = 1000 ◦ C 5.1 5.0 5.0 5.0 5.0

influenced by the variance of Zn, Co concentrations. The grains grow larger and porosity declines as the sintering temperature rises. Density data coincides with SEM results. Density increased with the rise of sintering temperature. The densities of the samples were determined by Archimedes’ method, and the results are listed in Table 2.

Fig. 4. Frequency spectra of the permeability in hyper-frequency for the CSAM samples with different compositions.

3.3. Magnetic properties of sintered samples Ferrite is ferrimagnetic material. The magnetization comes from the magnetic moment difference between the neighboring magnetic sublattices where spins align in antiparallel orientation due to superexchange interaction. Nonmagnetic Zn2+ ions preferentially occupy tetrahedral sites (A sites) where magnetic sublattices lie in antiparallel orientation to the whole lattice. Zn2+ ions in the sublattice do not change the spin directions; it simply reduces the magnetic moment of the Zn-containing sublattice. As a result, high Zn2+ content may lead to the increase of saturation magnetization. As Zn2+ ions are substituted by Co2+ ions, which have strong planar magnetic anisotropy, the saturation magnetization decreases and crystalline anisotropy increases. That determines permeability µ , according to the formula as follows: µ∝

Ms2 K1

Fig. 6 shows the frequency spectra of permeability of the sample by CSAM (x = 0.3) sintered at different temperatures. Permeability also increases with the rise of sintering temperature. This can be explained by µ=

µB − 1 µB (1 + δ/D) =1+ 1 + µB (δ/D) 1 + µB (δ/D)

(2)

where µ and µB stand for the effective permeability of the real samples and of the ideal bulk ferrite, respectively; D is the average size of the ferrite particles and δ is the average gap between them [8]. As it has been discussed above, grains grow larger and porosity declines as the sintering temperature rises. Hence, the gap parameter δ/D decreases with rise of sintering temperature. The effective permeability µ of the samples increases.

(1)

where µ stands for permeability, Ms stands for saturation magnetization, and K1 stands for magnetocrystalline anisotropy. As Co2+ amount increases, initial permeability reduces. The results follow the Snoek’s Law that permeability µ is in inverse proportion to cut-off frequency. Cut-off frequency rises with the increase of Co content. The frequency spectra of the permeability µ in hyper-frequency are shown in Fig. 4. For the x = 0 sample, a double resonant peak appears in the frequency spectrum. The resonance peak at low frequency originates from the magnetic domain wall resonance, whereas the one at high frequency from nature resonance. With the rise of Co content, the domain wall motion is blocked due to the enhancement of magnetocrystalline anisotropy. Hence, domain wall resonance disappears. At the same time, the nature resonance is driven to higher frequency [9]. With the same composition, the samples made by CSAM exhibit higher initial permeability than that of the samples made by SSRM (Fig. 5). The enhancement in permeability is more distinct in the samples with higher Zn contents.

Fig. 5. Frequency spectra of the permeability in hyper-frequency for the samples (x = 0.3–0.9) made by CSAM compared with those via SSRM. The CSAM samples were sintered at 900 ◦ C while SSRM samples were sintered at 1000 ◦ C.

70

Y. Bai et al. / Materials Chemistry and Physics 98 (2006) 66–70

tration. The cut-off frequency is in inverse proportion to permeability according to Snoek’s Law. (4) For the same composition, the samples made by SSRM exhibit higher permeability than those via CSAM. (5) Permeability increases with the rise of sintering temperature. Y-type hexagonal ferrites obtained by sol–gel autocombustion method have low sintering temperature (equal to or less than 900 ◦ C). It meets the need of LTCC technique to co-fire with Ag electrode without any addition. They exhibit excellent magnetic properties, such as high permeability and high cut-off frequency, in hyper-frequency. It is suitable to fabricate high quality multi-layer chip device for hyper-frequency. Acknowledgements Fig. 6. Frequency spectra of permeability of the sample (x = 0.3) prepared by CSAM and sintered at different temperatures.

4. Conclusions The important conclusions of our work can be summarized as follows: (1) The auto-combusted powders are nano-sized and highly active. After annealed at 900 ◦ C, pure Y-type hexagonal ferrite can be formed, while the phase formation temperature of Y-type phase by SSRM is 1000 ◦ C. (2) The samples fabricated by CSAM can be sintered well under 900 ◦ C without any addition, about 100 ◦ C lower than that by SSRM. (3) As Co2+ content rises, saturation magnetization decreases and magnetic planar anisotropy increases. That determines the decrease of permeability µ with rise of Co concen-

This work was supported by the Ministry of Sciences and Technology of China through 863-project under grant 2003AA32G030 and 973-project under grants of 2002CB61306 and 2001CB6104, and National Science Foundation of China under grants of 50425204, 50272032 and 90401012. References [1] [2] [3] [4] [5] [6]

Jen-Yan Hsu, IEEE Trans. Magn. 30 (1994) 4096. J. Smit, H.B.J. Wijn, Ferrites, Cleaver-Hume Press, London, 1959. Y. Bai, J. Zhou, Z. Gui, L. Li, J. Magn. Magn. Mater. 246 (2002) 140. S.G. Lee, S.J. Kwon, J. Magn. Magn. Mater. 153 (1996) 279. A. Nakono, Proc. ICF-6 (1995) 1225. Hongguo Zhang, Zhenwei Ma, Ji Zhou, Zhenxing Yue, Zhilun Gui, J. Magn. Magn. Mater. 213 (2000) 304. [7] Hongguo Zhang, Longtu Li, Ji Zhou, Jianer Bao, Zhenxing Yue, Zhilun Gui, J. Mater. Sci. -Mat. Elec. 11 (2000) 619. [8] Takanori Tsutaoka, Masahiro Ueshima, Toshihiko Tokunaga, Tatsya Nakamura, Kenichi Hatakeyama, J. Appl. Phys. 78 (1995) 3983. [9] Y. Bai, J. Zhou, Z. Gui, L. Li, Mater. Lett. 58 (2004) 1602.