Sintering temperature dependence of room temperature magnetic and dielectric properties of Co0.5Zn0.5Fe2O4 prepared using mechanically alloyed nanoparticles

ARTICLE IN PRESS Journal of Magnetism and Magnetic Materials 322 (2010) 686–691

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm

Sintering temperature dependence of room temperature magnetic and dielectric properties of Co0.5Zn0.5Fe2O4 prepared using mechanically alloyed nanoparticles Samaila Bawa Waje a,n, Mansor Hashim a,b, Wan Daud Wan Yusoff b, Zulkifly Abbas b a b

Advanced Materials and Nanotechnology Laboratory, Institute of Advanced Technology, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia Department of Physics, Faculty of Science, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

a r t i c l e in f o

a b s t r a c t

Article history: Received 14 June 2009 Received in revised form 14 September 2009 Available online 30 October 2009

Co0.5Zn0.5Fe2O4 nanoparticles were prepared using mechanical alloying (MA) and sintering. The crystallite size, coercivity, retentivity and saturation magnetization were also measured. The frequency dependence of dielectric and the magnetic parameters, namely, real permittivity e0 , loss tanget tan d, real permeability m0 and loss factor m00 were measured at room temperature for samples sintered from 600 to 1000 1C, in the frequency range 10 MHz to 1.0 GHz. The results show that the crystallite size of the resulting products ranges between 16 and 67 nm for as-milled sample and the sample sintered at 1000 1C, respectively. The sample sintered at 1000 1C, measured at room temperature exhibited a saturation magnetization of 37 emu g  1. The values of permittivity remain constant within the measured frequency, but vary with sintering temperature. The permeability values, on the other hand however vary with both the sintering temperature and the frequency, thus, the absolute value of the permeability decreased after the natural resonance frequency. & 2009 Elsevier B.V. All rights reserved.

Keywords: Mechanical alloying Sintering temperature Permeability Permittivity Nanoparticles

1. Introduction Magnetic nanoparticles have generated great interest due to their importance in fundamental understanding of physical processes and technological applications [1]. Due to the very small sizes of the particles involved, which are of the order of the magnetic domain sizes in the corresponding bulk materials, novel magnetic behaviors are observed for the nanosized magnetic particles when compared to that of the bulk counterparts [2]. The interesting magnetic properties of ferromagnetic spinel ferrites originate mainly from the magnetic interactions between cations with magnetic moments that are situated in the tetrahedral (A) and the octahedral (B) sites [3]. ZnFe2O4 is a normal spinel with all the Fe3 + ions in the B sites and all the Zn2 + ions in the A sites, whereas CoFe2O4 has an inverse spinel structure with the Co2 + ions mainly in the B sites and Fe3 + ions distributed almost equally between the A and the B sites [4]. Therefore, Co0.5Zn0.5Fe2O4 has a mixed spinel type of structure. Although the preparation of cobalt–zinc ferrites have been reported elsewhere [6–8], our literature search shows that there is no report on the permittivity and permeability within 10 MHz to 1.0 GHz reported, despite the materials potential use as a wave

n

Corresponding author. Tel.: + 60 389467546; fax: + 60 38656 6061. E-mail address: [email protected] (S.B. Waje).

0304-8853/$ - see front matter & 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jmmm.2009.10.041

absorber for electromagnetic interference (EMI). In this paper, we sintered mechanically alloyed (MA) nanoparticles to produce Co0.5Zn0.5Fe2O4 ferrite at a relatively low temperature, compared to conventional solid-state technique. The nanoparticles were firstly prepared by mechanical alloying and then sintered at various temperatures from 600 to 1000 1C in order to fabricate polycrystalline materials. Their microstructure properties are reported as a function of frequency and the sintering temperatures.

2. Materials and method All reagents used were of analytical grade from Alfar Aesar without any further purification. The starting powders of Co3O4 (99.7%), Fe2O3 (99.5%) and ZnO (99.0%) were weighed according to the targeted proportion and milled using a SPEX8000D in order to produce Co0.5Zn0.5Fe2O4 nanoparticles. Mechanical alloying was carried out at room temperature in a planetary ball mill equipped with a hardened steel vial and balls. In order to avoid an increase in the vial temperature, the milling procedure was interrupted for 5 min after every 15 min of milling. The milling media consisted of ten 12 mm diameter balls confined in a 120 ml volume container. A total of 8 g powder with no process control agent was used in all MA runs. The resulting particles were granulated using 2% PVA and lubricated using zinc stearate and then molded into samples of toroidal and disc shapes. The samples was

ARTICLE IN PRESS S.B. Waje et al. / Journal of Magnetism and Magnetic Materials 322 (2010) 686–691

sintered at various temperatures ranging from 600 to 1000 1C, for 6 h in each case using the heating rate of 21/min. The resulting products were finally used for surface, magnetic and dielectric measurements. Structural changes of powder particles were studied by X-ray diffraction (XRD) in a Philips X’PERT MPD diffractometer using filtered Cu Ka radiation (l = 0.1542 nm) obtained in the 2y range 20–701 using a scan step of (2y)= 0.03301 with 5 s per step as the counting. The average crystallite size of all the samples is determined from the full-width at half-maximum (FWHM) of the (3 1 1) reflection peak in the XRD patterns using the Debye Scherrer formula shown in Eq. (1). D¼

0:94 l b cos y

ð1Þ

where l is the wavelength of the incident X-ray; b the full-width at half-maxima; y is Bragg’s diffraction angle. An Accupyl 1330 Pycnometer was used to measure the density of the sample by measuring the pressure change of helium in a calibrated volume. Scanning electron microscopy (SEM) micrographs were taken using a Cambridge Stereoscan S-250 MK-II machine to reveal the microstructure of the resulting product. The specific saturation magnetization was measured at room temperature by a vibratingsample magnetometer (VSM) in a field of 10 kOe. The magnetic and dielectric measurements were carried out by the bridge method, using Agilent Model 4291B Network/Spectrum analyzer in the frequency range 10 MHz–1.0 GHz. The analyzer was equipped with HP16453A and HP16454A ‘L’ materials fixture for dielectric and magnetic measurements respectively. 3. Results and discussion The diffraction pattern of Co0.5Zn0.5Fe2O4 prepared at various temperatures is shown in Fig. 1 below. The signature peaks of the three starting raw materials, i.e Fe2O3 at 2y =33.181, ZnO at 2y =36.261 and Co3O4 at 2y =36.911 were evident for the sample before milling, and can be indexed to ICDD cards of 01-079-1741 for Fe2O3, 01-076-1802 for Co3O4 and 01-089-0511 for ZnO accordingly. For the as-milled powder, the XRD pattern shows peak intensity for the crystalline phase decreased consequent to the milling process, suggesting that the 12 h milling results in the

20000

687

introduction of some degree of amorphous-like phase, in addition to the traces of the 311 ferrite peak. This amorphous-like phase could not be identified from XRD profile, suggesting that the starting materials are completely alloyed, a similar phenomenon was observed elsewhere [9]. It is believed that the amorphouslike phase is a tetranary phase consisting of Co, Zn, Fe and O, which nucleates at interfaces and grows under interfacial metastable equilibrium. This is because; the ball milling facilitates such reactions by fracturing and cold-welding crystalline particles to create fresh interfaces, thereby generating a high density of defects. More so, the peaks became fairly weak and broadened, which indicates a reduction in the crystallite size. The lattice parameters of the prepared samples are shown in Table 1. For the as-milled sample, the ‘‘A’’ and ‘‘Vo’’ values showed a reduced cell volume and lattice parameter consequence to the alloying process. However, sintering resulted into a more perfectly regular packing of the spinel structure. The BET surface area and the average pore diameter of the as-milled sample are 35.34 m2/g and 18.50 nm, respectively. For the sintered samples, all the Co0.5Zn0.5Fe2O4 peaks appeared with the sintering temperature as low as 600 1C, indicating the degree of structural order (crystallinity). The XRD intensity counts, referred to the degree of crystallinity of the sample increased with increase in sintering temperature from 600 to 1000 1C. All the peaks were matched with the theoretically generated one with no peak being left un-indexed. The peaks can be indexed to (2 2 0), (3 1 1), (4 0 0), (4 2 2), (5 1 1) and (4 4 0) planes of a cubic unit cell, which correspond to cubic spinel structure. The indexed 2y values obtained have been compared with calculated values. A comparison of these values revealed good agreement between calculated and observed values as shown in Table 1. The crystallite size of the samples increased with an increase in the sintering temperature. The crystallite size of the samples shown in Table 1 below as calculated using the Scherrer equation ranges from 16 to 67 nm. The lattice parameter a for the sintered samples was found to be in the range ˚ which is within the range of the theoretical lattice 8.41–8.42 A, parameter values of 8.33 and 8.39 A˚ for zinc and cobalt ferrite, respectively. The observed deviation of the lattice parameter can be attributed to the rearrangement of cations in the nanosize Co0.5Zn0.5Fe2O4 crystallite consequent to the sintering process.

(311) (220)

(440)

(511)

(400)

(222)

(422)

1000 deg.

15000

900 deg. 800 deg.

10000

700 deg. 600 deg.

5000

As-milled (Fe2O3) (ZnO) (Co3O4) raw materials

0 20

30

40

50

60

Fig. 1. XRD micrographs of as-prepared samples of Co0.5Zn0.5Fe2O4.

70

ARTICLE IN PRESS 688

S.B. Waje et al. / Journal of Magnetism and Magnetic Materials 322 (2010) 686–691

Furthermore, the density of the samples as shown in Table 1 reveals an increase in the density with increase in sintering temperature, with the highest density of 5.43 g/cm3 for the sample sintered at 900 1C.

The morphology of the nanoparticles sintered at 900 1C as examined by SEM (Fig. 2a) shows agglomerated nanoparticles, which is attributed to the interaction between magnetic nanoparticles. The EDX spectrum in Fig. 2b shows clearly the

Table 1 Physical properties of the as-prepared Co0.5Zn0.5Fe2O4 samples. Sample

˚ A (A)

2y (Obs.)

2y (Cal.)

Vo (A˚ 3)

Crystallite size (nm)

Density r (g/cm3)

As-milled 600 1C 700 1C 800 1C 900 1C 1000 1C

7.998 8.411 8.418 8.420 8.416 8.418

35.458 35.341 35.393 35.368 35.363 35.349

35.464 35.348 35.370 35.321 35.344 35.335

345.04 593.79 594.89 595.87 596.10 596.52

16.(4) 38.(4) 44.(7) 53.(7) 67.(1) 67.(1)

5.1(3) 5.2(9) 5.3(3) 5.3(6) 5. 4(3) 5.3(8)

Fig. 2. Microstructures of the Co0.5Zn0.5Fe2O4 samples: (a) SEM, (b) EDX, (c) TEM of as-milled sample and (d) TEM of sample sintered at 900 1C.

ARTICLE IN PRESS S.B. Waje et al. / Journal of Magnetism and Magnetic Materials 322 (2010) 686–691

Zn, Co, Fe and O peaks, in addition to the Au peak, resulting due to the sample being sputtered with gold for conduction for clear viewing. The TEM images shown in Figs. 3a and b show the asmilled sample and the sample sintered at 1000 1C, respectively. The as-prepared sample shows a range of particles from 10 to

600°C

12

800°C

900°C

15 nm, while the sample (2d) ranges from 40 to 60 nm. Both results show some consistencies with the calculated values using the Scherrer formula. Figs. 4a and b shows the real permittivity (e0 ) and loss tangent (tan d) versus frequency plots, respectively. It is seen that all the

1000°C

600°C

0.45

800°C

900°C

1000°C

0.4

10

0.35 8

Loss Tangent

Permittivity Real

689

6 4

0.3 0.25 0.2 0.15 0.1

2

0.05 0 10

100 Frequency (MHz)

600°C

14

800°C

900°C

0

1000

10

1000°C

600°C

10

800°C

900°C

1000

1000°C

9

12

8 Loss Factor

10 8 6 4

7 6 5 4 3 2

2

1

0 10

100 Frequency (MHz)

0

1000

10

100 Frequency (MHz)

Fig. 3. Frequency dependence of (a) permittivity, (b) loss tanget, (c) permeability and (d) loss factor of the samples sintered at various temperatures.

as milled

sint. 600

sint 800

sint 900

sint. 1000

50 40 Mass Magnetization (emu/g)

Permeability Real

100 Frequency (MHz)

30 20 10 0 -15000

-10000

-5000

0

5000

10000

15000

-10 5 -20

3

-30

1

-40

-1 0

-400

-50 Field (G)

-200

200

-3 -5

Fig. 4. The VSM experimental results of Co0.5Zn0.5Fe2O4 sintered at various temperatures.

400

1000

ARTICLE IN PRESS 690

S.B. Waje et al. / Journal of Magnetism and Magnetic Materials 322 (2010) 686–691

Table 2 summary of the magnetic and dielectric properties of the as-prepared Co0.5Zn0.5Fe2O4. temp., 1C

As milled 600 800 900 1000

m0 , 1.0 GHz

m0 , 10 MHz

– 3.15 4.11 4.25 3.91

2.79 4.94 7.09 12.10

m00 ,

m00 , 10 MHz

e0 ,

e0 ,

1.0 GHz

1.0 GHz

10 MHz

tan d, 1.0 GHz

tan d, 10 MHz

Coercivity, (Oe)

Retentivity, (emu g  1)

Saturation at 10 kOe (emu g  1)

– 0.11 0.29 0.26 0.49

1.07 2.95 4.46 6.71

5.84 6.44 7.56 7.57

– 5.74 6.46 7.20 7.30

0.025 0.044 0.063 0.13

– 0.024 0.078 0.081 0.047

52 98 117 97 64

0.32 1.62 2.10 1.71 1.12

5..32 27.13 32.78 36.40 36.90

samples show dependence on sintering temperature, i.e. the dielectric constant increases with increase in sintering temperature. The observed higher values of e0 with sintering temperature might be due to the presence of a higher number of Fe2 + formed at elevated sintering temperatures [10,11]. However, no significant change was observed for the samples sintered at 900 and 1000 1C. Further observed is that all the samples show no significant frequency-dependent phenomenon, i.e. the permittivity maintains almost a constant value within 10 MHz–1.0 GHz (refer Table 2). For example, the e0 value for the sample sintered at 900 1C in the frequency range 10 MHz–1.0 GHz are 7.2 and 7.56, respectively, while the tan d values are around 0; precisely, they are 0.063 and 0.081, respectively. The fluctuations of the permittivity spectra are due to the extreme sensitiveness of the equipment which is within the acceptable measurement deviations ( 70.8%). There are two mechanisms responsible for the permittivity dispersion for polycrystalline ferrites at higher frequency; they are ion and electron polarization [12]. At high frequency, the e0 values are usually found invariable with frequency, and the tan d values are found very close to 0. Therefore, the dielectric losses for all samples are almost equal to 0. These results, as shown in Fig. 4 and Table 2 are consistent with well-accepted phenomenon. The room temperature initial permeability, m0 and loss factor 00 m values were measured from 10 MHz to 1.0 GHz for all the samples and the results are shown in Figs. 4c and d, respectively. From Fig. 4c, it can be seen that the m0 changes from 2.79 to 12.10 with sintering temperature (refer to Table 2). These according to Snoek’s relation given, m0 fr =constant, where fr represents the resonance frequency. It means that, the higher the values of m0 , the lower value of resonance. Thus, the higher m0 values at 10 MHz observed in the case of sample sintered at 1000 1C, resulted frequency in resonance at a lower frequency (200 MHz). The observed increase in m0 with sintering temperature can partly be attributed to the increased grain size of samples and also due to the hindrance to the spin movement, resulting thereby in the increased value of m0 [13]. Fig. 4d shows the loss factor m00 of the samples sintered at various temperatures. From the figure, it is observed that the losses remained constant with frequency from 10 to 120 MHz for samples sintered at 600–900 1C and around 80 MHz for the sample sintered at 1000 1C. Afterwards, the losses increase with increase in frequency. The variation in values with frequency showed a similar trend for all the samples. The values of m0 are known to depend on various factors such as stoichiometry, Fe2 + content and structural homogeneity, which in turn depends upon the composition and the sintering temperature of the samples [14]. The major contribution to the magnetic losses in ferrites is due to hysteresis losses, which is based on damping phenomena associated with irreversible wall displacement and spin rotations. However, the hysteresis loss becomes less important in the highfrequency range because the wall displacement is mainly damped and the hysteresis loss will be due to spin rotation [15]. The

sample sintered at 1000 1C shows higher loss at higher frequency, which is attributed to the higher permeability of the sample at lower frequency region, suggesting the suitability of the sample in EMI suppression within the measured frequency range. The obtain values of m0 and m00 at 10 MHz were higher than those reported elsewhere [16] for CoFe2O4, i.e E1.5 and 0.06, respectively. This increase is attributed to the magnetic interactions in ferrites. According to Neel’s postulate [17], magnetic moments of ferrites are a sum of magnetic moments of individual sublattices. Exchange interaction between electrons of ions in these sublattices has different value. Usually interaction between magnetic ions of sublattices ‘A’ and ‘B’ (A–B interaction) is the strongest. ‘A’–‘A’ interaction is almost ten times weaker and ‘B’–‘B’ interaction is the weakest. The dominant ‘A’–‘B’ interaction leads to complete or partial (non-compensated) ferrimagnetism [17]. Generally speaking, the value of magnetic moment in ‘‘B’’ lattice MB is much greater than the ‘‘A’’ lattice MA thus given rise to Eq. (2); Ms ¼ MB  MA 9

ð2Þ

Since ZnFe2O4 is a normal spinel, with the Zn2 + ions in the ‘A’ site and both Fe3 + ions in the ‘B’ site. On the other hand, CoFe2O4 is an inverse spinel with Co2 + in B site and Fe3 + distributed evenly between A and B sites. In CoFe2O4 ferrites one half of Fe3 + is placed in the A sites and another half in the B sites. The magnetic moments of Fe3 + are mutually compensated and the resulting moment of the ferrite is due to the magnetic moments of bivalent cations Co2 + in the B positions. Substitution of Co2 + with Zn2 + leads to introduction of non-magnetic Zn2 + ions into A sites, thus increasing the saturation magnetization, Ms, leading to an increased magnetization [18]. Fig. 4 and Table 2 show the changes in the hysteresis loops of the samples for both as-milled and the sintered samples at different temperatures. In all the sintered samples, a ferromagnetic behavior was observed. The saturation magnetization values were highly affected by increase in the temperature as a consequence of the gradual increase in the crystallinity and particle size, a similar behavior has been reported for other magnetic materials [19]. The samples relax back their spins by rotation on the removal of an applied magnetic field, giving a nearly zero net magnetic moment. The maximum magnetization was 36.90 emu/g and closely followed by 36.40 emu/g for samples sintered at 1000 and 900 1C, respectively. Both the corecivity and retentivity first increases from 52 Oe and 0.32 emu/g for the asmilled sample up to a maximum value of 117 Oe and 2.1 emu/g for the sample sintered at 800 1C; upon further annealing the particles at 1000 1C, it is observed that both the corecivity and retentivity decreases. This variation with crystallite size is also explained on the basics of domain structure, diameter of particles and crystal anisotropy. Since sintering temperature causes changes by decomposition or transformation of phases. These in turn brings about increase in grain size, change in pore shape, pore size and number. The starting materials, being in nanosize

ARTICLE IN PRESS S.B. Waje et al. / Journal of Magnetism and Magnetic Materials 322 (2010) 686–691

are not likely to have any grain boundary, however, sintering introduces the effects of grain boundary, in addition to some microstructural defects; thereby forming a source of flux B, thereby resulting to a higher values of corecivity and retentivity. Above 800 1C, the crystallite size increases, thereby decreasing the number of grain boundaries, decreasing the structural defects and subsequently decreased values of corecivity and retentivity.

4. Conclusion The present work shows preparation of cobalt–zinc ferrites prepared using mechanically alloyed nanoparticle. The desired cobalt ferrite cannot be formed directly through milling alone, thus, a heat treatment is necessary. This is because ball milling facilitates fracturing and cold-welding crystalline particles to create alternating layers with fresh interfaces. After an annealing of this powder at 600 1C, the spinel structure was obtained with a grain size in a nanometric range (40 nm). The as-prepared high density ferrite shows that the values of permittivity remains constant within the measured frequency, but however varies with the sintering temperature. On the other hand, the permeability values, however varies with both the sintering temperature and the frequency, and the absolute values of the permeability decreases after attaining the natural resonance frequency of the material. The as-prepared sample shows potential in electromagnetic wave suppression due to the high loss observed and therefore further study is in progress.

Acknowledgments The authors are grateful to Universiti Putra Malaysia for both Research University Grant (vote number 05-04-08-0548RU) and the Graduate Research Fellowship.

691

References [1] J.L. Dormann, D. Fiorani, in: Magnetic Properties of Fine Particles, NorthHolland, Amsterdam, 1992. [2] N.S. Gajbhiye, G. Balaji., Advances in nanoscience and nanotechnology, in: A. Sharma (Ed.), NISCAIR, 2003. [3] R.K. Sharmaa, O. Suwalkaa, N. Lakshmia, K. Venugopalana, A. Banerjeeb, P.A. Joyc, Synthesis of chromium substituted nano particles of cobalt zinc ferrites by coprecipitation, Mater. Lett. 59 (27) (2005) 3402–3405. [4] D.S. Mathew, R-S. Juang, An overview of the structure and magnetism of spinel ferrite nanoparticles and their synthesis in microemulsions, Chem. Eng. J. 129 (2007) 51–65. [6] M.M. Mallapura, P.A. Shaikha, R.C. Kambalea, H.V. Jamadara, P.U. Mahamunia, B.K. Chougule, Structural and electrical properties of nanocrystalline cobalt substituted nickel zinc ferrite, J. Alloys Compd. 479 (1–2) (2009) 797–802. [7] G. Vaidyanathana, S. Sendhilnathanb, Characterization of Co1  xZnxFe2O4 nanoparticles synthesized by co-precipitation method, Phys B: Condens. Matter 403 (13–16) (2008) 2157–2167. [8] M. Bahgat, Min-Kyu Paeka, Jong-Jin Paka, Comparative synthesize of nanocrystalline Fe–Ni and Fe–Ni–Co alloys during hydrogen reduction of NixCo1  xFe2O4, J. Alloys Compd. 466 (1–2) (2008) 59–66. [9] J. Ding, P. Cl., McCormick, R. Street, Magnetic properties of mechanically alloyed CoFe2O4, Solid State Commun. 95 (1) (1995) 31–33. [10] B. Parvatheeswara Rao, K.H. Rao, Effect of sintering conditions on resistivity and dielectric properties of Ni–Zn ferrites, J. Mater. Sci. 32 (22) (1997) 6049–6054. [11] H. Su, H. Zhang, et al., Complex permeability and permittivity spectra of polycrystalline Ni–Zn ferrite samples with different microstructures, J. Alloys Compd. 481 (1–2) (2009) 841–844. [12] J.B. BirksProgress in Dielectrics, 3, London Heywood & Company Ltd. pp. 153. [13] N. Gupta, Mukesh C. Dimri, Subhash C. Kashyap, Processing and properties of cobalt-substituted lithium ferrite in the GHz frequency range, Ceram. Int. 31 (1) (2005) 171–176. [14] A. Verma, T.C. Goel, R.G. Mendiratta, P. Kishan, Magnetic properties of nickel– zinc ferrites prepared by the citrate precursor method, J. Magn. Magn. Mater. 208 (2000) 13–19. [15] S. Chikazumi, in: Physics of Magnetism, Wiley, New York, 1964 263 pp. [16] S.B. Waje, M. Hashim, W.D. Wan Yusoff, Z. Abbas, Physical and magnetic characterization of polycrystalline cobalt ferrite (CoFe2O4) materials prepared via mechanically alloyed nanoparticles, J. Appl. Sci. Res. 5 (10) (2009) 1440–1444. [17] L. Neel, Magnetic properties of ferrites: ferrimagnetism and antiferromagnetism, Ann. Phys. Paris 3 (1948) 137–198. [18] H.S.C. O’Neill, A. Navrotsky, Simple spinels: crystallographic parameters, cation radii, lattice energies, and cation distribution, Am. Miner. 69 (1983) 181–194. [19] S. Morup., Superparamagnetism and spin glass ordering in magnetic nanocomposites, Europhys. Lett. 28 (1994) 671–676.