Journal of Alloys and Compounds 450 (2008) 387–394

Evolution of intermediate phases in the synthesis of zinc ferrite nanopowders prepared by the tartrate precursor method J.M. Yang a,∗ , F.S. Yen b a

Department of Industrial Engineering and Management, Diwan College of Management, Tainan 721, Taiwan, PR China b Department of Resources Engineering, National Cheng Kung University, Tainan 701, Taiwan, PR China Received 24 July 2006; received in revised form 17 October 2006; accepted 28 October 2006 Available online 1 December 2006

Abstract The advanced electronic applications of zinc ferrite (ZnFe2 O4 ) material are considered to require improvement in the powder processing, particularly, meticulous particle control in the nanometer range, stoichiometry and phase purity. This article presents the process of zinc ferrite formation from tartrate precursors, emphasizing the intermediate phase evolution during the thermal treatment. The variation of the intermediate phases, crystallite size and chemical composition of the products at different calcination temperatures have been investigated by X-ray diffraction (XRD), Fe2+ content analysis, transmission electron microscopy (TEM) and energy dispersive spectrometry (EDS) techniques. Results show that the stoichiometry and phase purity of the resultant zinc ferrite nanopowders has been noted to be highly influenced by their intermediate phases. Finally, the formation mechanism of zinc ferrite is discussed. A single phase, approaching the desirable stoichiometric zinc ferrite nanoparticles, is noted to be possibly prepared produced by an annealing treatment of the precursor gel powders at 350 ◦ C for 4 h. © 2006 Elsevier B.V. All rights reserved. Keywords: Spinel; Zinc ferrite; Tartrate; Nanoparticles; Stoichiometry

1. Introduction Ferrites are one of the most important classes of magnetic ceramic materials and are used extensively in electronic devices for the communications industry [1]. Recently, nanometersized magnetic ferrites have attracted considerable attention as improvements of their physical properties are achieved over that of the conventional feedstock [2–5]. Zinc ferrite (ZnFe2 O4 ) is of particular interest among ferrite spinels because of its use in providing a comprehensive scope for mixed ferrites meeting different specifications of magnetic materials [6–9], thereby, requiring a fine control over the size, crystallinity, stoichiometry and phase purity of the nanopowders. It may be mentioned that this study of ZnFe2 O4 represents an extension of our previous studies [10,11] of other ferrites, such as NiFe2 O4 and LiFe5 O8 . There have been various reports demonstrating the feasibility of wet chemical preparation of the ZnFe2 O4 nanoparticles via different approaches such as co-

∗

Corresponding author. Tel.: +886 6 571 8888x759; fax: +886 6 571 7146. E-mail addresses: [email protected], janne [email protected] (J.M. Yang). 0925-8388/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2006.10.139

precipitation [12], sol–gel [13], aerogel [14], hydrothermal [15], or decomposition of organometallic compounds [9,16–19], etc. These processes attempt to yield finer-sized powders by a more intimate mixing to the starting materials. The molecular precursor approach offers the advantage of achieving an intimate molecular-level mixing of the metal ions in the precursor. Tartrate complexation is one of the precursor methods for producing ZnFe2 O4 fine powders. This method requires a relatively low temperature and short duration to produce the oxide. On thermal conversion, the resultant precursor produces fine-particle oxide possessing desirable powder characteristics, such as uniform size and high surface area. However, most the pyrolytic decomposition of precursors procedures [9–11,16,18–22] have inadequacies in that the presence of unwanted secondary phases such as hematite (␣-Fe2 O3 ), giving delinquent influence on the stoichiometry and phase purity in the granular structure of the final product. It is further necessary to develop preparation methods, which improve the particle stoichiometry, enhancing the crystallinity and phase purity at a relatively low calcination temperature. It seems that a systematic study of the nature of progress of thermal reaction and phase evolution is necessary for understanding the important factors for the determinations of

388

J.M. Yang, F.S. Yen / Journal of Alloys and Compounds 450 (2008) 387–394 Fluka, 99.5%) was dissolved in ethyl alcohol (Seoul, 99.5%) to produce a tartaric acid solution, and (ii) based on an atomic ratio of Zn/Fe = 1/2, reagentgrade zinc acetate (Zn(CH3 COO)2 ·2H2 O, Merck, 99.9%) and ferric nitrate (Fe(NO)3 ·9H2 O, Merck, 99.9%) were dissolved in ethyl alcohol and then stirred at room temperature until the solution became transparent. The cation solution was then added to the tartaric acid solution. The resulting solution precipitated at room temperature over the course of 1 h as it was stirred constantly to ensure the completion of the complex reactions. The solution was then dried at 80 ◦ C for 24 h. The dried gel was ground to −200 mesh ( 0.4 range, the chemical composition of the ZnFe2 O4 is very close to that of stoichiometric ZnFe2 O4 . Note that these observations are consistent with the XRD phase evolutions as shown in Fig. 3. The compositional analysis of the particles (Fig. 5) confirms that the spinel-type Zn-ferrite product formed at lower

Fig. 5. Histogram of particle chemical composition distribution (Zn/Fe atomic ratio) obtained by EDS for samples calcined at temperatures of: (a) 350 ◦ C, (b) 400 ◦ C, and (c) 600 ◦ C.

temperatures (325–400 ◦ C) is not stochichiometric, and results in ␣-Fe2 O3 precipitation at higher temperatures (>500 ◦ C). Therefore, it is apparent that the formation of ␣-Fe2 O3 depends strongly on the composition of the Zn-ferrites (nonstoichiometric versus stoichiometric), which is formed at a lower temperature, and has a significant effect on the phase purity of the final product. 3.5. Summary of characteristics and phase evolution of calcined products The summary of characteristics and phase evolution of the products calcined from 275 ◦ C to 600 ◦ C in four separate stages is given in Table 1. (i) The calcined products at calcinations temperature from 275 ◦ C to 600 ◦ C are noted to have different colors, namely: from black, brown, orange, brownish red and dark-red, with the order of increasing temperatures. (ii) Significantly, there is a prominent increase of crystallite sizes for different specimens heated from 275 ◦ C to 600 ◦ C, namely: from 5 nm to 45 nm, respectively.

J.M. Yang, F.S. Yen / Journal of Alloys and Compounds 450 (2008) 387–394

391

Fig. 6. TEM dark fields of: (a) ␥-Fe2O3 and (b) ␣-Fe2O3. Note that ED patterns and EDS analysis results indicate presence of substantial amounts of Zn in ␥- and ␣-crystallites.

(iii) The sequence of phase evolution is noted as the the following: amorphous, ZnO, Fe3 O4 , ␥-Fe2 O3 , non-stoichiometric Zn-ferrite, stoichiometric Zn-ferrite and ␣-Fe2 O3 for specimens calcined from 275 ◦ C to 600 ◦ C, respectively. (iv) Mole ratio of Fe2+ /Fetotal may be used as an index for discriminating the particular phases. The mole ratio for products of the above sequence (iii) varies: from 0.51 ± 0.01 to 0.00 ± 0.01, respectively.

and Fe3 O4 phase by the decomposition of the Zn–Fe tartrate precursor according to the reaction:

These results above indicate that Zn-ferrite is formed according to a multi-step process. The first step is the formation of ZnO

The second step is the oxidation of Fe3 O4 to ␥-Fe2 O3 , which is accompanied by the creation of cation vacancies as shown in

Zn1 Fe2 -tartrate precursor 270−300 ◦ C

300−325 ◦ C

−−−−−→amorphous−−−−−→ 23 Fe3 O4(s) + ZnO(s)

(1)

392

J.M. Yang, F.S. Yen / Journal of Alloys and Compounds 450 (2008) 387–394

Table 1 Summary of phase evolution and characteristics in calcined powders as determined by X-ray diffraction and Fe2+ contents analysis Thermal reaction

Characteristics (◦ C)

Phases obtained

Fe2+ /Fetotal

5±2

Amorphous

0.51 ± 0.01

– – 10 ± 0.1 –

Amorphous ZnO Fe3 O4 ZnO

0.32 ± 0.01

Brown

11 ± 0.1 – 10 ± 0.1 –

Fe3 O4 ZnO ␥-Fe2 O3 NZFa

0.24 ± 0.01

400

Orange Brown-red

± ± ± ±

␥-Fe2 O3 NZF ZFb ␣-Fe2 O3

0.08 ± 0.01

550

12 16 32 37

600

Dark-red

ZF ␣-Fe2 O3

0.00 ± 0.01

Stage

Calcined temperatures

Ix

275

Black

IIx1

300

Black

325

Black

IIx2

350

IIIx

IVx a b

Phase evolution

Color

Crystallite size (nm)

0.1 0.2 2 3

45 ± 5 53 ± 6

0.31 ± 0.01

0.00 ± 0.01

NZF: non-stoichiometric Zn-ferrite. ZF: ZnFe2 O4 .

the following equation: 2 3 Fe3 O4(s)

325−400 ◦ C

+ 16 O2(g) −−−−−→ 43 Fe8/3 1/3 O4(s)

(2)

The third step is the formation of Zn-ferrite in a solid-state reaction between ZnO and ␥-Fe2 O3 , that is 325−450 ◦ C

␥ − Fe2 O3(s) + xZn2+ −−−−−→Znx Fe(8−2x)/3 (1−x)/3 O4(s) (3.1) 325−450 ◦ C

Znx Fe(8−2x)/3 (1−x)/3 O4(s) −−−−−→Zn(x+y) Fe(8−2(x+y))/3 (1−(x+y))/3 O4(s)

(3.2)

precursors at lower processing temperatures (typically below 500 ◦ C) and in a very short reaction time. Obviously, the atomic diffusion is slower at a lower temperature range and the atomic rearrangement, occurring in a regime for long-range diffusion, is constrained [32]. Therefore, in the present synthesis of Znferrite, if the temperature at each reaction step is not maintained for a sufficient length of time, Zn2+ diffusion into the ␥-Fe2 O3 structure (Eq. (3.1)) may fail to complete, causing the composition of the ZnFe2 O4 to deviate markedly from the ideal ZnFe2 O4 stoichiometry and to have a lower Zn/Fe ratio (Fig. 5). Therefore, when synthesizing ZnFe2 O4 , it is essential to specify a sufficiently long calcination period to ensure the full diffusion of Zn2+ into the ␥-Fe2 O3 structure.

In this equation, 0  x, y  1, and x, y depends on the extent of the reaction. The general formula, Znx Fe(8−2x)/3 (1−x)/3 O4 , is that of NZF, which can be considered as solid solutions between the ZnFe2 O4 and the defective spinel ␥-Fe2 O3 , respectively. The thermal stability field of the Znx Fe(8−2x)/3 (1−x)/3 O4 is dependent on the Zn content (x). Finally, further heating the NZF (Znx Fe(8−2x)/3 (1−x)/3 O4 or Zn(x−y) Fe(8−2(x−y))/3 (1−(x−y))/3 O4 ) at higher temperatures (>500 ◦ C), the thermodynamically stable phases ␣-Fe2 O3 and ZnFe2 O4 are obtained by heating, in accordance with the following reaction: Znx Fe(8−2x)/3 (1−x)/3 O4(s) >500◦ C

−−−−→xZnFe2 O4(s) + 4(1 − x)/3α − Fe2 O3(s)

(4)

3.6. Effect of annealing for ZnFe2 O4 formation Kinetically, inorganic compounds, as the ceramic materials, are generally synthesized by the pyrolytic decomposition of

Fig. 7. XRD patterns of products obtained by: (a) annealing at 350 ◦ C for 4 h and (b) annealing at 650 ◦ C for 2 h.

J.M. Yang, F.S. Yen / Journal of Alloys and Compounds 450 (2008) 387–394

393

a TEM micrograph of the ZnFe2 O4 particles in the sample annealed at 350 ◦ C for 4 h. The particle diameter is estimated to be approximately 20 nm. Figs. 7 and 8 confirm that this annealing treatment produces a desirable single phase and approaching an ideal stoichiometric Zn-ferrite nanopowder. 4. Conclusion

Fig. 8. Histogram of Zn/Fe atomic ratio distribution obtained by EDS analysis by TEM of sample annealed at 350 ◦ C for 4 h.

The starting gel powders were calcined at 350 ◦ C and then annealed for 4 h. Fig. 7 demonstrates the effect of heat treatment during the synthesis of Zn-ferrite. The XRD pattern shows that the annealed specimen have a spinel structure (Fig. 7a). When the annealed sample (Fig. 7a) was further calcined at 650 ◦ C for 2 h, the XRD pattern in Fig. 7b shows that the crystal structure remained unchanged. The only observable effect of the additional annealing is that the degree of X-ray line broadening decreases as a result of the growth of larger crystallites at the higher treatment temperature. Significantly, Fig. 7b reveals that the ␣-Fe2 O3 phase is eliminated. EDS chemical analyses show that the composition distribution of the sample annealed at 350 ◦ C for 4 h coincides almost exactly with the stoichiometric value of ZnFe2 O4 (Zn/Fe = 0.50) (Fig. 8). Fig. 9 presents

This study has employed a tartrate precursor approach to synthesize ultrafine particles of Zn-ferrite. The results show that the desired stoichiometric ZnFe2 O4 is not produced directly by the thermal decomposition of the Zn–Fe tartrate precursor. Prior to the formation of ZnFe2 O4 a series of intermediate phases, ZnO, Fe3 O4 , ␥-Fe2 O3 and NZF (Znx Fe(8−2x)/3 (1−x)/3 O4 ), are identified. Specifically, the EDS analyses have shown that the Znferrite nanoparticles formed at low temperatures (325–450 ◦ C) are NZF. This leads to the precipitation of ␣-Fe2 O3 at higher calcined temperatures (>500 ◦ C). On the other hand, for the precursor gel powders annealed at 350 ◦ C for 4 h, a single phase and approaching an ideal stoichiometric Zn-ferrite nanopowder can be obtained, which may possibly give beneficial electronic/magnetic properties. Acknowledgments We wish to thank Dr. H.S. Liu and Y.H. Chang for their valuable helpful discussions. We also wish acknowledge to C.M. Chen and L.J. Wang for their assistance in performing the current XRD and TEM analyses. References

Fig. 9. TEM micrograph of sample obtained for the precursor gel powders annealed at 350 ◦ C for 4 h.

[1] U. H¨afeli, W. Sch¨utt, J. Teller, M. Zborowski (Eds.), Scientific and Clinical Applications of Magnetic Carriers, Plenum Press, New York, 1997. [2] T. Sato, T. Iijima, M. Seki, N. Inagaki, J. Magn. Mater. 65 (1987) 252–256. [3] K. Tanaka, M. Makita, Y. Shimizugawa, K. Hirao, N. Soga, J. Phys. Chem. Solids 59 (1998) 1611–1618. [4] C.N. Chinnasamy, A. Narayanasamy, N. Ponpandian, K. Chattopadhyay, Mater. Sci. Eng. A 304–306 (2001) 983–987. [5] T. Sato, K. Haneda, M. Seki, T. Iijima, Appl. Phys. A 50 (1990) 13–16. [6] W. Schiessl, W. Potzel, H. Karzel, M. Steiner, G.M. Kalvins, M. Martin, K. Krause, I. Halevy, J. Gal, W. Sch¨afer, W. Will, M. Hillberg, R. W¨appling, Phys. Rev. B 53 (1996) 9143–9152. [7] B. Gillot, M. El Guendouzi, J. Solid State Chem. 106 (1993) 443–450. [8] K. Suresh, K.C. Pati, J. Mater. Sci. Lett. 14 (1995) 1074–1077. [9] V. Moye, K.S. Rane, V.N.K. Dalal, J. Mater. Sci.: Mater. Elect. 1 (1990) 212–218. [10] J.M. Yang, W.J. Tsuo, F.S. Yen, J. Solid State Chem. 145 (1999) 50–57. [11] J.M. Yang, W.J. Tsuo, F.S. Yen, J. Solid State Chem. 160 (2001) 100–107. [12] B. Jeyadevan, K. Tohji, K. Nakatsuka, J. Appl. Phys. 76 (1994) 6325–6327. [13] C.D.E. Lakeman, D.A. Payne, Mater. Chem. Phys. 38 (1994) 305–324. [14] H.H. Hamdeh, J.C. Ho, S.A. Oliver, R.G. Willey, G. Oliveri, G. Busca, J. Appl. Phys. 81 (1997) 1851–1857. [15] A. Clearfiled, A.M. Gadalla, W.H. Marlow, T.W. Livingston, J. Am. Ceram. Soc. 72 (1989) 1789–1792. [16] P. Ravindranathan, K.C. Patil, Am. Ceram. Soc. Bull. 66 (1987) 688–692. [17] N.S. Gajbhiye, U. Bhattacharya, V.S. Darshane, Thermochim. Acta 264 (1995) 219–230. [18] F.J. Guaita, H. Beltran, E. Cordoncillo, J.B. Carda, P. Escribano, J. Eur. Ceram. Soc. 19 (1999) 363–372.

394

J.M. Yang, F.S. Yen / Journal of Alloys and Compounds 450 (2008) 387–394

[19] B.S. Randhawa, J. Mater. Chem. 10 (2000) 2847–2852. [20] R.J. Willey, S.A. Oliver, G. Oliveri, G. Busca, J. Mater. Res. 8 (1993) 1418–1427. [21] Y. Hayashi, T. Kimura, T. Yamaguchi, J. Mater. Sci. 21 (1986) 2876–2880. [22] I. Mohai, J. Sz´epv¨olgyi, I. Bert´oti, M. Mohai, J. Gubicza, T. Ung´ar, Solid State Ionics 141–142 (2001) 163–168. [23] B.D. Cullity, Elements of X-ray Diffraction, Addison-Wesley, Reading, MA, 1978, 284. [24] I. Vogel, A Textbook of Quantitative Inorganic Analysis, Longmans, New York, 1961, 308.

[25] B. Gillot, R.M. Benloucif, A. Rousset, J. Solid State Chem. 39 (1981) 329–336. [26] B. Gillot, R.M. Benloucif, Mater. Chem. Phys. 32 (1992) 37–41. [27] G. Dupre, A. Rousset, P. Mollard, Mater. Res. Bull. 11 (1976) 473–476. [28] B. Gillot, A.R. Rousset, G. Dupre, J. Solid State Chem. 25 (1978) 263–271. [29] P.S. Sidhu, R.J. Gilkes, A.M. Posner, J. Inorg. Nucl. Chem. 39 (1977) 1953–1958. [30] W. Feitknecht, K.J. Gallagher, Nature 228 (1970) 548–549. [31] T. Sano, Y. Tamaura, Mater. Res. Bull. 34 (1999) 389–401. [32] C.G. Levi, Acta Mater. 46 (1998) 787–800.