Mechanosynthesis of Gadolinium Iron Garnet S. C. Zanatta, L. F. Co´tica, A. Paesano, Jr., w and S. N. de Medeiros Departamento de Fı´sica, Universidade Estadual de Maringa´ , PR, Brazil J. B. M. da Cunha Instituto de Fı´sica, Universidade Federal do Rio Grande do Sul, RS, Brazil B. Hallouche Departamento de Quı´mica e Fı´sica, Universidade de Santa Cruz do Sul, RS, Brazil We investigated the mechanosynthesis of gadolinium iron garnet by high-energy ball milling stoichiometry amounts of a-Fe 2 O 3 and Gd 2 O 3 , followed by short thermal annealings conducted at moderate temperatures. X-ray diffraction and Mo¨ ssbauer spec- troscopy results revealed, for the as-milled samples, the forma- tion of the GdFeO 3 perovskite phase, in relative amounts that depend on the milling time. The formation of the Gd 3 Fe 5 O 12 garnet phase was observed in 10001C/2 h or 11001C/3 h as-an- nealed samples. The occurrence of a milling time was verified in which the relative amount of garnet phase formed by further annealing is maximized. I. Introduction T HE gadolinium iron garnet (GdIG) has been extensively in- vestigated because of its technological significance, particu- larly regarding microwave applications. 1–3 The importance of gadolinium-substituted YIG (yttrium iron garnet), the most widely used rare-earth iron garnet, 4,5 is the existence of a region in the magnetization versus temperature curve, between the compensation point, T CP , and the Curie temperature, T C , in which the magnetization is approximately constant. dM/dT  0 makes the difference for some particular microwave devices. 6 The process conventionally conducted to synthesize this ce- ramic phase is the solid-state reaction (SSR) among Fe 2 O 3 and Gd 2 O 3 powders, mixed in stoichiometry amounts of 5:3. 7,8 This method requires a calcination temperature of over 14001C and this high temperature inevitably leads to coarsened microstruc- tures, which affect the sinterability of the synthesized pow- ders. 9,10 When obtained at lower temperatures, a similar garnet (i.e., Gd 3 Fe 3 Al 2 O 3 ) demanded a prolonged heating of 2 weeks. 1 Certainly, this has implications for the production costs of garnets. Aiming to overcome these inconveniences, some alternative routes for garnet preparation such as co-precipitation, sol–gel, hydrothermal synthesis, spray, and freeze-drying methods have been investigated. 9–12 On the other hand, high-energy ball milling (HEBM) has been successfully used to mechanosynthesize a wide range of ceramic powders (see, e.g., Suryanarayana 13 and references therein). This mechanical process often enhances the reaction of multi-component systems by significantly lowering the tem- perature for single-phase formation. The yttrium aluminum gar- net (YAG), for instance, was obtained at a temperature significantly lower after being activated by HEBM. 9,14 In this sense, we have investigated the formation of GdIG by high-energy ball milling followed by heat treatments conducted at moderate annealing conditions, characterizing the samples in each stage of preparation by X-ray diffraction (XRD) and, mainly, by Mo¨ ssbauer spectroscopy (MS). The latter is partic- ularly suitable for characterization of iron garnets because in- formation about the iron’s local chemical environment, charge, and magnetic state may be obtained by using the 57 Fe nuclear probe. The phases and structure formed (i.e., nano or bulk structured) may also be identified from the hyperfine parame- ters’ isomer shift (d), quadrupole splitting (QS), and hyperfine magnetic field (B hf ). II. Experimental Procedure The samples were prepared, firstly, by high-energy ball milling analytical-grade a-Fe 2 O 3 and Gd 2 O 3 oxides, manually pre- mixed in nominal compositions of 5:3. The precursors were dry milled for times ranging between 1 and 24 h, in an argon atmosphere, in a Fritsch Pulverisette 6 planetary ball mill (Idar- Oberstein, Germany), using a hardened steel 80 cm 3 vial charged with hardened 10 mm diameter steel balls. The vial was not opened during the milling process. The ball-to-powder mass ra- tio (20:1) and rotation speed of the supporting disk and vial (300 rpm) were kept constant throughout the experiments. Subse- quent to the milling, the samples were annealed in free atmos- phere either at 10001C for 2 h or at 11001C for 3 h. The XRD patterns of the as-milled and annealed products were obtained at room temperature (RT), using a Siemens D500 X-ray diffractometer (Munich, Germany) in Bragg–Brentano geometry, with CuKa radiation, in the 201r2yr701 range. The MS characterization was performed in the transmission geom- etry, using a conventional Mo¨ ssbauer spectrometer in a constant acceleration mode. The g-rays were provided by a 57 Co(Rh) source. The Mo¨ ssbauer spectra were analyzed with a non-linear least-squares routine, with Lorentzian line shapes. Occasionally, a hyperfine magnetic field distribution (histogram), B hf Dist, was used in the spectral analysis. The values of isomer shift are referred to that of an a-Fe foil at 300 K. III. Results and Discussion Figure 1 shows the XRD patterns for some selected as-milled samples. The diffractograms show no clear phase formation un- til 3 h of milling (Fig. 1(b)), although the reflection peaks for a- Fe 2 O 3 and Gd 2 O 3 are greatly broadened and their intensities are strongly reduced as a consequence of the grain size refinement and of very disordered or defected structures. J ournal J. Am. Ceram. Soc., 88 [12] 3316–3321 (2005) DOI: 10.1111/j.1551-2916.2005.00598.x r 2005 The American Ceramic Society 3316 L. Levinson—contributing editor Supported by the Brazilian agencies: CAPES (PROCAD), CNPq, and Funda@a˜ o Arauca´ ria. w Author to whom correspondence should be addressed. e-mail: paesano@wnet.com.br Manuscript No. 20285. Received March 8, 2005; approved May 18, 2005.