Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
Magnetic and magnetotransport properties of nanocrystalline (Fe2B)0.20X0.80 (X=Ag or Cu) alloys prepared by mechanical alloying E.C. Passamania,*, J.R.B. Tagarroa, E. Nunesa, A.Y. Takeuchib a
Departamento de F!ısica, Universidade Federal do Esp!ırito Santo, Campus de Goiabeiras, Av. Fernando Ferrari S/N, ! 29060-900 Vitoria, ES, Brazil b Centro Brasileiro de Pesquisas F!ısicas, R. Dr. Xavier Sigaud 150, 22290-180, Rio de Janeiro, RJ, Brazil Received 13 November 2001; received in revised form 13 February 2002
Abstract The magnetic and magnetotransport properties of nanocrystalline (Fe2B)0.20X0.80 (X=Ag or Cu) alloys have been . studied by Mossbauer, magnetization and magnetoresistance (MR) techniques. The samples were prepared by mechanical alloying the nanostructured Fe2B alloy and Ag or Cu chemical elemental powders in a high-impact mill . machine. The Mossbauer spectroscopy and magnetic measurements of as-milled (Fe2B)0.20Ag0.80 and annealed (Fe2B)0.20Cu0.80 alloys indicate the presence of small Fe and Fe2B magnetic particles. The magnetoresistance curves of (Fe2B)0.20X0.80 (X=Ag or Cu) alloys show a non-saturated behavior, an indication of a spin-glass-like state generated by the magnetic coupling between these magnetic particles. The non-saturated effect is more pronounced in the alloy with Ag, where the MR values decrease linearly with the external magnetic field, even at 7 T. The magnetoresistance values, at 4.2 K and for a magnetic field of 7 T, are 4.0% and 4.5%, for (Fe2B)0.20Ag0.80 and (Fe2B)0.20Cu0.80 samples, respectively. r 2002 Elsevier Science B.V. All rights reserved. Keywords: Nanocrystalline materials; Single magnetic domains; Magnetic precipitates; Spin-glass-like state; Magnetoresistance
1. Introduction The discovery of giant magnetoresistance (GMR) in artificially structured systems has attracted much interest in the past few years because of their potential applications. The GMR effect was first found in Fe–Cr multilayers by Baibich et al. [1] and it was later reported that the
*Corresponding author. Fax: +55-27-335-2823. E-mail address:
[email protected],
[email protected] (E.C. Passamani).
GMR can also be found in systems with a positive enthalpy, which often form small magnetic particles dispersed in a non-magnetic matrix. Berkowitz et al. [2] and Xiao et al. [3] observed for the first time GMR effects in granular magnetic systems when studying metastable films of Fe–Cu and Cu–Co prepared by sputtering deposition, followed by heat treatments. Wecker et al. [4] used the melt spinning method and produced ribbons of metastable CoCu alloys: subsequent heat treatment leads to the formation of Co particles in a Cu matrix, having a GMR value of 36% at 30 K and H=6 T.
0304-8853/02/$ - see front matter r 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 8 8 5 3 ( 0 2 ) 0 0 1 5 7 - 9
192
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
The mechanical alloying (MA) induced by highimpact milling (HIM) is another technique considered capable of forming supersaturated or granular alloys from immiscible systems. Therefore, the MA method can produce metastable alloys that may consist of metallic magnetic particles in a non-magnetic matrix and, consequently, GMR materials. The great advantage of MA is the easy formation in large amounts of bulk materials by compacting powders under pressure. The GMR effect has been studied in several binary alloys produced by HIM, such as Co–Ag [5], Fe–Cu [6], Co–Cu [7] and Fe–Ag [8]. Although the GMR behavior has been observed in these types of immiscible systems, the magnetoresistance (MR) of ternary systems, with Cu and Ag used as a non-magnetic matrix, has been seldom reported. The advantage of producing ternary alloys is the possibility to change the magnetic properties of the magnetic precipitates. Therefore, the ternary alloys are suitable systems to study binary alloy precipitates, directly obtained either by milling through the reduction of the alloy particle sizes or by annealing the formed supersaturated ternary alloys. The binary metastable alloys produced by HIM, using Cu and Ag as a solid solvent show different results concerning the final crystalline structure. For the Ag systems [8], the granular features are obtained directly by the milling procedures, while for the Cu ones the supersaturated binary alloys are formed first and then the small magnetic precipitates are obtained by annealing, at specific temperatures that are usually determined by thermal measurements [6,7]. One expects that on applying MA to Fe2B-X (X=Ag or Cu) systems, the formed alloy will have a granular structure when Ag is used as a solid solvent and that there will be a supersaturated phase in the Cu case. Therefore, annealing at certain temperatures will produce precipitation of Fe or Fe2B small particles for the Cu system, while these small particles can be probably formed directly by milling in the case of the Ag matrix. So, one expects to observe the effects of such microstructure on the MR measurements in both systems. In the present paper, we report a study of the magnetic and magnetotransport properties of mechanically alloyed (Fe2B)0.20X0.80 (X=Ag or
Cu) powders. The non-magnetic matrix composition of 80 at% was chosen in order to inhibit the rapid growth and the magnetic contact among the magnetic particles obtained during the milling process or after heat treatment.
2. Experimental (Fe2B)0.20X0.80 (X=Ag or Cu) alloys were obtained from high-purity chemical elemental powders of Fe, B and Cu or Ag (99.999%). The milling process was initiated with a small amount of the mixture of the elemental powders, at the defined composition, in order to cover the cylinder (milling tool) and walls of the vial, therefore, reducing the contamination from them. A larger amount of the powder mixture was then introduced and the milling process was performed with a massive cylinder which occupies 63% of total internal space of the vial. Before starting the milling, a few drops of methanol was added as a process control agent, an experimental procedure that has been used to promote grain refinement, i.e. to avoid a cold welding of the powders during the milling process. The samples were prepared in two steps: first by milling the elemental powders of Fe and B for 310 h, at the composition of the Fe0.67B0.33 alloy, here labeled as Fe2B. In a second step, Ag or Cu powder was added at the defined composition and the mixtures were milled for 217 and 150 h, respectively. The manipulation of samples was done inside a glove box under highpurity Ar atmosphere to prevent the oxidation and contamination with other gases. The total mass of each final product was about 10 g. Part of the final (Fe2B)0.20Cu0.80 alloy was sealed in a quartz tube under vacuum, and heat treated at 873 K for 24 h. This temperature was selected from differential scanning calorimetry (DSC) analysis, that shows an exothermic peak at about 850 K. X-ray diffraction (XRD) patterns for all the samples were collected at 300 K using Cu-Ka radiation in a Rigaku diffractometer, that was calibrated using a standard graphite crystal. 57Fe . Mossbauer spectra for all the samples were measured with a 1-year-old 50 mCi 57Co:Rh radioactive source in a conventional transmission
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
3. Results and discussion 3.1. (Fe2B)0.20Ag0.80 alloy
30
40
50
Fe2 B Fe 2B
Fe 2B
Relative amplitude
Fe 2B
Fe
60
70
Fe
80
90
100
2θ (Degree)
Fig. 1. X-ray diffraction pattern of the Fe2B alloy milled for 310 h.
217 h
150 h
Relative amplitude
. geometry at room temperature (RT). A Mossbauer spectrum at 10 K of the (Fe2B)0.20X0.80 alloy milled for 150 h was obtained in a closed cycle He . refrigerator, with the Mossbauer source at RT. . Each Mossbauer spectrum of the milled or annealed samples were collected during one week. The isomer shift (IS) values of our data were taken relatively to a-Fe at RT. Magnetic hysteresis loops (M vs. H curve) of (Fe2B)0.20X0.80 (X=Ag (as-milled) or Cu (annealed)) alloys were examined at 4.2 K and at RT, using a vibrating-sample magnetometer. The MR measurements were carried out at 4.2 K using a four-point technique with magnetic fields up to 7 T. The samples [(Fe2B)0.20X0.80 (X=Ag (asmilled) or Cu (annealed))] for the MR measurements were cold pressed into pellets at 1.8 kbar and at 300 K, for 10 min. The final compositions of the alloys were checked by electron scanning microscopy using energy dispersive X-ray spectroscopy (EDS) analysis. The EDS results show that there were no traceable impurities in the final milled samples.
193
125 h
50 h
In this paper we will not focus the attention on the Fe2B system prepared by MA, but it is worth mentioning that the Fe2B milled alloy is composed of Fe2B (89%) and Fe (11%) phases, similarly to the results reported by Yang et al. [9]. The XRD pattern of the Fe2B milled for 310 h shows the presence of these two nanocrystalline phases (see Fig. 1). The characterization of magnetic properties of the Fe2B alloy produced by 310 h of milling will be published elsewhere. Fig. 2 displays the XRD patterns of the (Fe2B)0.20Ag0.80 alloy milled at the indicated times. The XRD patterns of the alloy show that the (2 0 0), (2 2 0) and (2 2 2) Bragg peaks of FCC-Ag are very close to the most intense peaks of Fe2B and BCC-Fe, respectively. Therefore, after a short milling time, it is not possible to distinguish the Fe and Fe2B X-ray diffraction peaks from those mentioned Bragg peaks of FCC-Ag phase, may
3h
30
40
50
60
70
80
90
100
2θ θ (Degree)
Fig. 2. X-ray diffraction patterns of the (Fe2B)0.20Ag0.80 powder milled at different times. Lozenge, square and circle symbols correspond to Ag, Fe2B and Fe phases, respectively.
be due to line broadening caused by the small crystalline grain sizes of these phases. However, the most intense (1 1 1) Bragg peak of FCC-Ag has a different angular position and, therefore, can be used to study the alloy formation. Fitting the (1 1 1) FCC Bragg peak, for different milling times, with a Lorentzian line, one can observe that: (a) its angular position does not change with the increase
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
194
in the milling time and (b) the line width increases with the milling process. The unchanged angular position of the FCC-Ag peak may be an indication of no alloy formation between Ag and the Fe2B and Fe phases, while the line width broadening effect is associated with a FCC-Ag grain reduction, a characteristic of crystalline materials produced by HIM down to nanocrystalline sizes. . Fig. 3 shows the RT Mossbauer spectra for powder samples of (Fe2B)0.20Ag0.80 alloy milled for 102, 170 and 217 h. The original two sextets of
0.6%
170 h
102 h
0.2%
Relative Transmission
1%
217 h
-8
-6
-4
-2
0
2
4
6
8
Velocity (mm/s) . Fig. 3. RT Mossbauer spectra of the (Fe2B)0.20Ag0.80 powder milled at different times.
the non-milled alloy (310 h Fe2B sample) associated with a-Fe with Bhf ¼ 33 T and IS=0.0 mm/s (S1) and Fe2B with /Bhf S=23 T and /ISS= 0.18 mm/s (S2) phases are transformed into two other sextets (‘‘S1’’ and ‘‘S2’’), respectively, and one doublet. The corresponding hyperfine parameters of those subspectra used to fit the . Mossbauer spectra of the (Fe2B)0.20Ag0.80 alloy milled for 102, 170 and 217 h are shown in Table 1. . From Fig. 3, it can be observed that all Mossbauer spectra show broad lines, a characteristic of granular systems at nanometric scale. These two sextets ‘‘S1’’ and ‘‘S2’’, obtained after the milling process, have their magnetic hyperfine field (Bhf ) values slightly smaller than the ones measured for the non-milled alloy, i.e., S1—Fe (33 T) and S2—Fe2B (23 T). The ‘‘S1’’ sextet has hyperfine parameters that are close to those observed by Gome! z et al. [8] for Fe–Ag milled alloys and its presence is attributed to a fraction of Fe atoms forming small Fe particles distributed on the surface or at the interface of the Ag grains. On the other hand, the attribution by Gome! z et al. [8] for these Fe particles is based on: (a) the M vs. H curve of the final milled material which displays a hysteresis loop with a coercive field (HC ) of 328 Oe, while HC is close to zero for the non. milled alloy and (b) the Mossbauer spectra were composed of two magnetic sextets: one with a-Fe hyperfine parameters and a second one with slightly smaller Bhf value. The second magnetic component was attributed to Fe particles. Using the same approach of Gome! z et al. [8], one may attribute the ‘‘S2’’ sextet to small Fe2B particles, since this component has the same IS value of the Fe2B phase, but smaller Bhf (22 T). Therefore, the
Table 1 . . Mossbauer hyperfine parameters obtained from the fittings of the Mossbauer spectra of the (Fe2B)0.20Ag0.80 milled sample. The IS values are referred to as a-Fe at RT Milling time (h)
102 170 217
Fe2B on the Ag surface
Fe on the Ag surface
Fe–Ag phase
IS (mm/s)
Bhf ðTÞ
G (mm/s)
! AREA (%)
IS (mm/s)
Bhf ðTÞ
G (mm/s)
! AREA (%)
IS (mm/s)
QS (mm/s)
G (mm/s)
! AREA (%)
0.06 0.05 0.06
32.4 32.3 32.4
0.40 0.30 0.30
24 31 31
0.21 0.21 0.22
23.0 21.8 21.8
0.28 0.28 0.28
46 45 46
0.26 0.26 0.27
0.57 0.56 0.57
0.32 0.32 0.32
30 24 23
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
magnetic subspectra of MA (Fe2B)0.20Ag0.80 powders, in fact, have two magnetic components: one attributed to the small Fe particles and the other one due to small Fe2B particles, both dispersed on the surface or at the interface of Ag grains. This model of small Fe2B and Fe particles is also supported by the magnetization measurements, where non-zero HC values are measured for the final milled alloy at 4.2 and 300 K. . Fig. 3 also shows that the Mossbauer spectra of (Fe2B)0.20Ag0.80 MA powders contain a nonmagnetic subspectrum (doublet). By comparing the hyperfine parameters (Table 1) of the nonmagnetic component with data on Fe–Ag alloys reported in the literature [10,11,12], one may assign the non-magnetic component to a diluted
195
Fe–Ag phase. However, this doublet is broader than those reported, an effect which may be associated with a contribution of Fe2B or Fe phases in superparamagnetic state, which make undistinguishable the dilute Fe–Ag component from the small particles of the Fe2B or Fe phases. The occurrence of a superparamagnetic behavior of the Fe2B and Fe particles may be understood by the temperature behavior of HC which will be discussed below. Fig. 4 shows the magnetization curves measured at 300 K (a) and 4.2 K (b) for (Fe2B)0.20Ag0.80 MA powders milled for 217 h. For comparison, the curve for the non-milled sample is also included, showing a negligible value of HC (Fig. 4(c)). From this figure, it can be observed that the
12
H C =373 Oe
8
M (emu/g)
4 0 -4 -8
(a)
-1 2 -1 2
-10
-8
-6
-4
-2
0
2
4
6
8
10
12
Applied magnetic field (kOe) 12
H c = 957 Oe
M (emu/g)
8 4 0 -4 -8
(b)
-1 2 -1 2
-10
-8
-6
-4
-2
0
2
4
6
8
10
12
Applied magnetic field(kOe) 12
M (emu/g)
8 4 0 -4 -8 -12 -12
(c) -10
-8
-6
-4
-2
0
2
4
6
8
10
12
Applied magneticfield (kOe) Fig. 4. Magnetization curves recorded at 4.2 K (a) and 300 K (b) for the (Fe2B)0.20Ag0.80 powder milled for 217 h. The magnetization curve (c) measured at 300 K for the initial (Fe2B)0.20Ag0.80 powders is also shown for comparison.
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
(Fe2B)0.20Ag0.80 MA powders exhibit hysteresis loops at both temperatures. The HC value obtained at RT from the M vs. H measurement is in the range of the values observed in singledomain magnetic particles in conventional magnetic materials [13]. The HC value for the (Fe2B)0.20Ag0.80 sample, milled for 217 h, increases about 170%, while the magnetization, at the highest applied magnetic field, increases only about 14%, from 4.2–300 K. This unusual HC behavior with temperature may be associated with a blocking effect on magnetic particles that freeze the magnetic moment of the superparamagnetic particles at low temperatures. Thus, from the HC results one may conclude that the (Fe2B)0.20Ag0.80 alloy milled for 217 h has a distribution of particle sizes, since at RT the M vs. H curve displays a non-zero HC value which increases strongly at low temperatures. The granular behavior is also confirmed by the MR measurements presented below. Fig. 5 shows the result of the MR (Dr=r0 ) measurements at 4.2 K for the (Fe2B)0.20Ag0.80 MA powders. Dr=r0 is defined as Dr=r0 ¼ ðrðHÞ rðH ¼ 0ÞÞ=rðH ¼ 0Þ and the Dr=r0 data were fitted with a straight line, giving the following function: Dr=r0 ¼ 0:53ð4ÞH þ 0:1ð1Þ; where H is the applied field in Tesla and the numbers in parentheses are the errors from the fitting. The fitting procedure does not correspond to a physical model, but only give us an idea for values of Dr=r0 at different applied fields. We notice that Dr=r0 decreases by 4.0% in a field of 7 T, and displays a non-saturated linear behavior. This dependence
0
has also been discussed by Sumiyama et al. [14] ! and Gomez et al. [8], for Fe–Ag alloys and it is assumed to be due to the strong magnetic interactions between the small Fe particles dispersed on the surface or at the interface of Ag grains, yielding a spin-glass-like state. Therefore, the non-saturated linear behavior of the MR curve observed at 4.2 K for (Fe2B)0.20Ag0.80 cold pressed sample is associated with the magnetic interactions of small Fe and Fe2B particles distributed on the surface or at the interface of the Ag grains. It should be pointed out, however, that our MR value of 4.0% at 7 T is slightly higher than that observed for the Fe–Ag milled alloy (at about 2%), at similar Fe composition and measured at 4.2 K up to 8 T [8]. 3.2. (Fe2B)0.20Cu0.80 alloy 3.2.1. Results for the (Fe2B)0.20Cu0.80 alloy for different milling times Fig. 6 displays the XRD patterns of the (Fe2B)0.20Cu0.80 alloy milled for different times,
150 h
100 h
Relative amplitude
196
68 h
-1
-3 α− Fe
∆ρ / ρ0 (% )
10 h -2
0h
-4
-5 0
1
2
3
4
5
6
7
8
Applied magnetic field (T)
Fig. 5. MR curve at 4.2 K of the milled (Fe2B)0.20Ag0.80 powders cold pressed at 300 K.
30
40
50
60
70
80
90
100
2θ (Degree)
Fig. 6. X-ray diffraction patterns of the (Fe2B)0.20Cu0.80 powder milled at different times.
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
as indicated. The XRD pattern of the non-milled sample shows Bragg peaks of FCC-Cu and BCCFe phases. The apparent absence of the Fe2B phase in the XRD pattern of the non-milled sample can be understood due to the similarity in the angular positions of its main peak with that of the FCC-Cu phase. For the milled sample it was observed that after a short milling time (Tm E10 h) the BCC-Fe peaks are no longer detected in the XRD patterns, indicating that the Fe atoms have diffused into Cu matrix forming a supersaturated alloy. The angular positions of the FCC-Cu peaks shift slightly to lower angles with increasing milling times. This shifting effect observed for FCC-Cu peaks is characteristic of Fe–Cu alloys and is associated with a substitution of Fe atoms in Cu matrix [15]. . The room-temperature Mossbauer spectra for the (Fe2B)0.20Cu0.80 milled alloy have the same paramagnetic component for milling times longer than 68 h, indicating that the process of alloy formation has already been achieved at that time. . Fig. 7 shows the Mossbauer spectra measured at 10 K (a) and at 300 K (b) of the (Fe2B)0.20Cu0.80 . alloy milled for 150 h. Both Mossbauer spectra shown in this figure are formed by a paramagnetic doublet. At 300 K, the hyperfine parameters of this component are IS of 0.20 mm/s and QS of 0.50 mm/s. In Fig. 7, the second-order Doppler
1%
Relative Transmission
(a)
(b)
-8
-6
-4
-2
0
2
4
6
8
Velocity (mm/s)
. Fig. 7. Mossbauer spectra of the (Fe2B)0.20Cu0.80 powder milled for 150 h and recorded at 10 K (a) and 300 K (b). The vertical dotted line displays the contribution of second-order . Doppler shifting for the Mossbauer spectra.
197
shift contribution for the IS value can also be observed, since in the low temperature measurement source and absorber are at different temperatures. Therefore, by combining the X-ray and . Mossbauer results collected at 10 K it can be concluded that the (Fe2B)0.20Cu0.80 sample forms a supersaturated alloy for milling times longer than 68 h, being paramagnetic even at 10 K. On the other hand, the M vs. H curve for the (Fe2B)0.20Cu0.80 alloy milled for 150 h measured at 300 K (not shown here) displays a magnetic hysteresis loop with a coercive field of 280 Oe, a characteristic of conventional magnetic granular materials [13]. Since this magnetic phase, observed by magnetic measurements, is not defined in the . Mossbauer spectra and that the XRD patterns shown only an FCC phase for milling times longer than 68 h, we suggest that this magnetic phase may be associated with small Fe2B magnetic particles. On the other hand, the apparent absence of this . phase in Mossbauer spectra can be understood as . due to the large background line (low Mossbauer signal), an effect already discussed in the previous work on the Fe–Co–Cu milled alloy [15]. 3.3. Results for the (Fe2B)0.20Cu0.80 alloy milled for 150 h (as-milled) and annealed at 873 K for 24 h The XRD pattern for the (Fe2B)0.20Cu0.80 alloy milled for 150 h and annealed at 873 K for 24 h is shown in Fig. 8(b). The results of the as-milled (Fe2B)0.20Cu0.80 alloy (150 h) is also displayed for comparison (Fig. 8(a)). It is observed that for the annealed sample the FCC Bragg peaks shift slightly towards the Cu angular positions, and there is also the presence of the most intense Bragg peak for the BCC-Fe phase. Thus, the evidence of BCC-Fe in the XRD pattern of the annealed sample is indicative of the precipitation of small BCC-Fe particles in the FeBCu alloy matrix. The assumption of a FeBCu matrix, instead of a Cu . one, is supported by the Mossbauer spectrum of the annealed sample, which has the same paramagnetic doublet observed in the as-milled (Fe2B)0.20Cu0.80 alloy (see Fig. 9(a) as-milled and (b) annealed at 873 K for 24 h). This result for the annealed sample indicates that boron atoms entering into Fe–Cu system favor the thermal
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
198
20
Relative Amplitude
15
Hc = 373 Oe
10 M (emu/g)
(a)
5 0 -5
α-Fe
-10 -15
(b)
-20 -12
30
35
40
45
50 55 60 2θ (Degree)
65
70
75
-9
80
Fig. 8. X-ray diffraction patterns of the (Fe2B)0.20Cu0.80 powder milled for 150 h: (a) as-milled and (b) annealed. Full circle and the star symbols in the annealed sample correspond to FeBCu and Graphite phases, respectively. Graphite powder . was used to fill the Mossbauer sample holder.
-6 3 -3 0 6 Amplitude magnetic field (kOe)
9
12
Fig. 10. Magnetization curve measured at 4.2 K for the (Fe2B)0.20Cu0.80 alloy milled for 150 h.
0
∆ρ / ρ0 (% )
-1
Relative Transmission
1%
(a)
-2
-3
-4
-5 0
1
2
3
4
5
6
7
Applied magnetic field (T)
Fig. 11. MR curve at 4.2 K of the 150 h milled and annealed (Fe2B)0.20Cu0.80 powders cold pressed at 300 K.
0.3%
(b)
-12
-8
-4
0
4
8
12
Velocity (mm/s)
. Fig. 9. RT Mossbauer spectra of the (Fe2B)0.20Cu0.80 powder milled for 150 h: (a) as-milled and (b) annealed.
. stabilization of the alloy. Again, the Mossbauer spectrum of the annealed sample is very poor and no magnetic contribution can be distinguished in it. Fig. 10 displays the M vs. H curve obtained at 300 K for the (Fe2B)0.20Cu0.80 alloy milled for 150 h and annealed at 873 K for 2 h. This curve is similar to the one observed for the untreated sample, although with slightly larger HC values.
This may be due to the small Fe particles also observed in the XRD pattern (Fig. 8 (b)). Therefore, due to the indicative presence of magnetic particles of Fe and Fe2B in the annealed (Fe2B)0.20Cu0.80 alloy, in a similar fashion as in the Ag-based system, an MR measurement at 4.2 K and for applied fields up to 7 T was performed. The obtained curve is shown in Fig. 11, where Dr=r0 is defined as before for the Ag system. However, in this case the shape of the Dr=r0 curve for the annealed (Fe2B)0.20Cu0.80 alloy is different from the one found for the (Fe2B)0.20Ag0.80 alloy. In the case of the Cu system, it can be observed that a relative abrupt
E.C. Passamani et al. / Journal of Magnetism and Magnetic Materials 247 (2002) 191–199
reduction of the electrical resistivity for applied fields up to 2 T, while for higher fields the MR curve seems to show also a non-saturated feature as observed in the Ag system. The behavior of MR curves for fields up to 2 T may be an indication that the small magnetic precipitates are not in a percolation regime in the Cu case, as compared with those magnetic particles in Ag system that yield a spin-glass-like state. In the Cu case, Dr=r0 decreases by 4.5% in a field of 7 T, similar to the results reported for the Fe–Cu milled alloys [6]. The non-saturated trend in the MR curve at higher applied magnetic fields observed for the annealed (Fe2B)0.20Cu0.80 sample may be understood in a similar way as for the Ag system, i.e., it may be due to a magnetic interaction between small Fe and Fe2B precipitates forming a spin-glass-like state.
4. Conclusions The magnetic and magnetotransport properties of the mechanically alloyed (Fe2B)0.20X0.80 (X=Ag or Cu) cold-pressed powders were studied. The presence of small Fe and Fe2B magnetic particles in these samples is demonstrated by the . Mossbauer spectroscopy and magnetization techniques. The MR measurements show a nonsaturated behavior, even under high magnetic fields, indicating a probable presence of small interacting Fe and Fe2B particles. This suggests a spin-glass-like state, as shown by the low-temperature measurements. The MR values at 4.2 K for an applied magnetic field of 7 T are 4.0 and 4.5% for the (Fe2B)0.20Ag0.80 and (Fe2B)0.20Cu0.80 powders, respectively. From the behavior of MR measurements of the two systems (Cu and Ag), it can be suggested that the small magnetic particles do not reach a magnetic percolation regime in Cu case, as compared with those particles in Ag system, since in the Cu case the MR curve has first an abrupt reduction of electric resistance for fields of about 2 T.
199
Acknowledgements The authors thank Prof. K.M.B. Alves for the nice discussions, Dr. S. Garcia for critical reading of the manuscript and the technician Paulo Cesar Martins da Cruz for the X-ray measurements. Also, we would like to acknowledge the financial support given by UFES, CNPq and the PCI/MCT program.
References [1] M.N. Baibich, J.M. Broto, A. Fert, F. Nguyen van Dan, F. Petroff, P. Etienne, G. Creuzet, A. Friederich, J. Chazeles, Phys. Rev. Lett. 61 (1988) 2472. [2] A.E. Berkowitz, J.R. Mitchell, M.J. Carey, A.P. Young, S. Zhang, F.E. Spada, F.T. Parker, A. Hutten, G. Thomas, Phys. Rev. Lett. 68 (1992) 3745. [3] J.Q. Xiao, J.S. Jiang, C.G. Chien, Phys. Rev. Lett. 68 (1992) 3749. [4] J. Wecker, R. von Helmolt, L. Schuz, K. Samwer, Appl. Phys. Lett. 62 (1993) 1985. [5] J.M.D. Coey, A.J. Fagan, R. Skomski, J. Gregg, K. Ounadjela, S.M. Thompson, IEEE Trans. Magn. 30 (1994) 666. [6] S.S. Saxena, J. Tang, Y.S. Lee, C.J. D’ Connor, J. Appl. Phys. 76 (1994) 6820. [7] A.Y. Yermakov, M.A. Uimin, A.V. Shangarov, A.V. Zarubin, Y.V. Chechetkin, A.K. Shtolg, V.V. Kondratyev, G.N. Konygin, Y.P. Yelsukov, S. Enzo, P.P. Macri, R. Frattini, N. Cowlam, Mater. Sci. Forum 225–227 (1996) 147. [8] J. Gom!ez, S.K. Xia, E.C. Passamani, B. Giordanengo, E. Baggio-Saitovitch, J. Magn. Magn. Mater. 223 (2001) 112. [9] Y.Z. Yang, Q.S. Li, Y.L. Zhu, X.M. Ma, Y.D. Dong, J. Mater. Technol. 15 (1999) 137. [10] C.L. Chien, K.M. Unruh, Phys. Rev. B 28 (1983) 1214. [11] Y. Nakamura, K. Sumiyama, N. Kataoka, Hyper. Inter. 28 (1986) 1029. [12] C. Larica, E. Baggio-Saitovitch, S.K. Xia, J. Magn. Magn. Mater. 110 (1992) 106. [13] B.D. Cullity, Introduction to Magnetic Materials, Addison-Wesley Publishing Company, New York, 1972. [14] K. Sumiyama, S.A. Makhlouf, S. Yamamura, Y.F. Xu, T.J. Konno, K. Wakoh, T. Hihara, K. Suzuki, J. Magn. Magn. Mater. 140–144 (1995) 563. [15] V.P. Nascimento, E.C. Passamani, A.Y. Takeuchi, C. Larica, E. Nunes, J. Phys.: Condens. Matter. 13 (2001) 665.
