* Corresponding author. Tel.: #55-21-542-2804, fax: #55- 21-541-7657. E-mail address: cozac@epq.ime.eb.br (J.C. Neto). Journal of Magnetism and Magnetic Materials 226}230 (2001) 727}729 Detailed investigation of the GMI in a commercial soft-ferromagnetic amorphous alloy Joa  o Cozac Neto*, F.L.A. Machado, R.S. de Biasi, A.E.P. de Araujo, F.M. de Aguiar Departamento de Engenharia Mecanica e de Materiais, Praca General Tiburcio 80, Urca, 22290-270, Rio de Janeiro, RJ, Brazil Departamento de Fn & sica, Universidade Federal de Pernambuco, 50670-901 Recife, PE, Brazil Abstract An investigation of the giant magnetoimpedance e!ect (GMI) in pieces of a commercial alloy METGLAS 2705 M is presented. The GMI measurements were carried out in as-quenched and annealed samples. The composition C of the samples was investigated by energy-dispersive X-ray spectrometer (EDS) and ferromagnetic resonance has been measured in the X-band (9.5 GHz). It is found that C departs from the nominal value when the electron beam is swept from the center of the ribbon to the edge. On the other hand, while the annealing increases the GMI it decreases the FMR linewidth. Despite the inhomogeneity in the composition it was also found that the GMI is larger in wider sam- ples.  2001 Elsevier Science B.V. All rights reserved. Keywords: Giant magnetoimpedance; Ferromagnetic resonance; Amorphous systems*ribbons; Amorphous systems*soft magnet The giant magnetoimpedance (GMI) e!ect in soft- ferromagnetic amorphous alloys has attracted much at- tention in the scienti"c community lately due to its po- tential use in the development of new magnetic sensor technologies. The e!ect is so called because the impe- dance varies by orders of magnitude when small mag- netic "elds (H) are applied. The GMI is essentially due to the fact that the surface magnetoimpedance is inversely proportional to the skin depth, , which in turn, is a func- tion of the magnetic permeability [1}3]. Investigation of the GMI in commercial materials is particularly interest- ing because they are readily available for technological applications. The alloy, of nominal composition Co  Fe  Ni  Mo  B  Si  , was supplied in the form of ribbons 25 mm wide and 20 m thick. Isothermal heat treatments were carried out in air on samples with typical dimensions of 2.5cm2mm20 m cut from the as- quenched ribbons. First-derivative FMR spectra were recorded at room temperature using an X-band ( f"9.5 GHz) Varian E-12 spectrometer. All measure- ments were taken with H parallel to the sample surface along the long axis of the ribbon. GMI measurements were carried out using a four-probe technique. The con- tacts were prepared using silver paint. The GMI ratio was de"ned as Z/Z ("10[Z(H)!Z(H  )]/Z(H  )), where H  ("3200 A/m) is the maximum value of the H, obtained from a Helmholtz coil. The amplitude of the electrical current was varied in the range 6}24mA. In order to study the e!ect of annealing, the temper- ature (¹  ) was increased in small steps and the GMI measurements were performed before and after each step. The results are shown in Fig. 1 for a frequency of 100 kHz and a current of 20 mA for samples cut from the central region of the ribbon. The maximum GMI e!ect was observed for ¹  "363 K. At these annealing conditions one obtains much larger values for the GMI than those obtained under "eld treatment [4]. The increase in the GMI e!ect is attributed to the structural relaxation, which improves the magnetic properties. Indeed, we con- "rmed that the samples are more homogeneous after annealing by measuring the FMR in a sample before and 0304-8853/01/$- see front matter  2001 Elsevier Science B.V. All rights reserved. PII:S0304-8853(00)01151-3