Magnetic and Structural Properties of Fe Mn Al Alloys Produced by Mechanical Alloying

Transcription

1 Hyperfine Interactions 148/149: , Kluwer Academic Publishers. Printed in the Netherlands. 295 Magnetic and Structural Properties of Fe Mn Al Alloys Produced by Mechanical Alloying G. A. PÉREZ ALCÁZAR 1, C. GONZÁLEZ 1 and B. CRUZ 2 1 Departamento de Física, Universidad del Valle, A.A , Cali, Colombia 2 Departamento de Física, Universidad Tecnológica de Pereira, La Julita, A.A. 097, Pereira, Colombia Abstract. The effects of concentration and preparation conditions on the magnetic and structural properties of samples of the Fe Mn Al system prepared by mechanical alloying were studied by Mössbauer spectroscopy, X-ray diffraction and susceptibility techniques. X-ray diffraction showed that samples with high Al content (40 at.%) and prepared with low milling times (lest than 12 hours) present the peaks of Mn, Fe Al alloy and broad peaks which correspond to the ternary bcc alloy. For 12 hours the Mn peak disappears and for 24 h or more the bcc ternary phase is totally formed. For t 24 h Mössbauer spectroscopy studies proved that: samples rich in Fe are ferromagnetic (F) at room temperature (RT) and present a reentrant spin-glass (RSG) behavior for low T ; samples with intermediate Fe values are paramagnetic (P) at RT and SG for low T ; and those with low Fe content are P at RT, antiferromagnetic (AF) at low T and RSG at very low T. By ac magnetic susceptibility it was possible to detect an additional magnetic phase in the F and P regions which was attributed to a superparamagnetic (SP)-like behavior. XRD and Mössbauer spectroscopy of samples with low Al content (7.5 at.%) showed that they present a fcc AF phase and a Fe Al F phase for low milling times ( 10 h). For bigger times the only detected phase is the fcc, which is AF for low Fe contents and P for high Fe contents. The effect of using small balls is to enhance the alloyed process. Key words: mechanical alloying, 57 Fe Mössbauer spectrometry, X-ray diffraction. 1. Introduction Disordered Fe Mn Al alloys present an enormous experimental and theoretical scientific interest due to the different magnetic phases experimentally detected. In general these alloys were prepared following the Chakrabarti method [1]. In the bcc structural phase magnetic phases such as paramagnetic (P), ferromagnetic (F), antiferromagnetic (AF), spin-glass (SG) and reentrant spin-glass (RSG) in the F and AF phases [2 4], respectively, were detected. Recently the pure and reentrant super paramagnetic (SP) phases were experimentally postulated [5, 6] in this alloy system and big theoretical effort is now in progress in order to interpret this behavior. The presence of all these magnetic phases is due to the competitive interaction of the Fe and Mn atoms which behave as F and AF, respectively, and to the disordered character of these alloys. In the fcc structural phase only two magnetic phases were reported at RT, the AF for low Al and Fe contents and the P phase for high Al or

2 296 G. A. PÉREZ ALCÁZAR ET AL. Fe contents [7, 8]. In this phase the AF character is due to Mn atoms and the P character is due to Al and Fe atoms. For low T the AF phase is the stable one for binary Fe Mn alloys [9] and ternary Fe Mn Al alloys [10]. The aim of the present work is to present the more recent experimental results on the magnetic and structural properties of different samples series of this system prepared by mechanical alloying and with different milling conditions. 2. Experimental procedure Samples series of Fe x Mn 0.6 x Al 0.4,0.2 x 0.6, were mechanically alloyed using the 15 : 1 ball mass to powder mass (BM/PM) ratio and different milling times from 2 to 48 hours. Ball milling experiments were carried out with a planetary ball mill with vials and balls of hardened chromium steel, using Ar atmosphere and pure (>99.9%) element powders. Samples series of Fe x Mn x Al C 0.01 Cu 0.003, 0.50 x 0.75, were milled using ratios of 4 : 1 and 12 : 1 at different times and two ball sizes, 1 and 2 cm. X-ray diffractograms (XRD) were taken for different samples. The fit of the peaks, in order to obtain the lattice parameters and grain sizes, were performed using Lorentzians. The grain size was calculated employing the Scherrer formula [11]. Mössbauer spectra were realized in a conventional spectrometer with an α-fe foil as the calibration sample. Some powder samples were also pressed with cylindrical shape of 3 mm diameter and 3 mm height for ac magnetic susceptibility, under an ac field of amplitude T, a biasing dc field of T and at a frequency of 175 Hz. Milled samples with 7.5 at.% Al were sieved at different meshes. 3. Experimental results and analysis 3.1. COMPOSITION AND MILLING TIME EFFECT IN SAMPLES WITH HIGH Al CONTENT Samples of the Fe x Mn 0.6 x Al 0.4 series were milled in order to test the effect of composition and the milling time. Figure 1 shows the XRD patterns, for 2 between 40 and 47 degrees, of samples with x = 0.35andafter2,4,6,8,12,24and 48 hours of milling. It can be noted that for 2 and 4 hours only two peaks appear, one corresponds to the principal peak of α-mn and the other was fitted with the (1 1 0) peak of the Fe and the (1 1 1) peak of the Al. For 6 and 8 hours a fourth peak between the previous ones was necessary for an adequate fit. This peak corresponds to the (1 1 0) of the bcc Fe Mn Al disordered alloy [3]. For 12 hours the Mn peak disappears, that of Fe Mn Al alloy increases and those of the Fe and Al are coinciding. Finally, for 24 and 48 hours only the ternary alloy peak appears showing that for 24 hours this alloy is totally consolidate. The XRD patterns obtained for samples with other x values are very similar, but with a decreasing Mn peak as x increases. Similar results for samples with 20 and 30 at.% Al were reported [12].

3 MAGNETIC AND STRUCTURAL PROPERTIES OF Fe Mn Al 297 Figure 1. XRD patterns of samples with x = 0.35 for different milling times. Figure 2 shows the obtained mean grain size of the bcc phase as a function of the milling time for the x = 0.35 sample. It can be noted that the mean grain size increases, decreases, and increases from 6 to 24 hours and then slowly decreases to 120 Å for 48 hours. This result is a little different of that reported for similar alloys but with lest Al content [12], in which the grain size increases continuously up to a stable value. The present result is the typical behavior expected for brittle systems which, in the present case, is induced by the high Al content. Similar results were obtained for samples with other x values and 40 at.% Al (not showed here). For samples with x = 0.35 the lattice parameter vs. milling time of the bcc phase is nearly constant with a value of 2.93 Å. Similar results were obtained for samples with other x values, but with a lattice parameter which decreases with the increases of the Fe concentration. This fact is a consequence of the smaller atomic size of the Fe atoms in relation to that of the Mn atoms. Figure 3 shows the Mössbauer spectra (MS) at room temperature (RT) of the x = 0.35 samples milled at different times. The spectra for 2 and 4 hours were fitted with a hyperfine field distribution and a broad singlet showing the disordered

4 298 G. A. PÉREZ ALCÁZAR ET AL. Figure 2. Mean grain size vs. milling time for the x = 0.35 samples. Figure 3. Mössbauer spectra at RT of samples with x = 0.35 for different milling times.

5 MAGNETIC AND STRUCTURAL PROPERTIES OF Fe Mn Al 299 character of the sample. The HFD can be attributed to the detected Fe which presents some Al atoms as next nearest neighbors, very similar to the reported behavior for melted Fe Al with low Al content [13]. The broad singlet can be attributed to Fe atoms inside the detected Al lattice and showing Fe sites rich in Al atoms (paramagnetic sites). For other milling times the P part of the spectra, which was fitted with a doublet and a singlet, increases and that of the F site decreases as the milling time increases in such way that for 24 hours the ferromagnetic contribution disappears. The P part of the spectra were fitted in two ways, one using a single line and a doublet and other by using a single line and quadrupolar distribution in order to attend the disordered character of the samples. These two methods give very good fits but the second one gives a very narrow quadrupolar distribution which can be assumed as only one doublet. Then we adopt the first fitting method. The fitted Mössbauer parameters for the spectra of the samples milled during 24 and 48 hours are δ = 0.20 ± 0.02 mm/s for the singlet and δ = 0.18 ± 0.02 mm/s and QS = 0.42 ± 0.02 mm/s for the doublet. Then, in according with XRD results, the consolidate bcc ternary phase is P and disordered with some Fe sites surrounded by a symmetric charge distribution (singlet) and other Fe sites surrounded by an asymmetric charge distribution (doublet). Similar results were obtained for samples with other x values, but when x increases the consolidate alloy tends to be ferromagnetic as is illustrated in Figure 4. In this figure, typical spectra for some samples milled during 48 hours are showed. At RT the x 0.40 samples are P and for x>0.40 they are F. Figure 5 shows the ac susceptibility vs. temperature curves obtained for the consolidated samples (milled during 48 hours) with different x values. It can be noted a peak at 280 K in the x = 0.20 sample, a change of curvature at 50 K and a peak at 210 K in the x = 0.25 sample, a broad peak which begin at 50 K and finish at 105 K and a broad peak at 240 K in the x = 0.30 sample, a peak at 50 K in the x = 0.35 sample, a peak at 50 K and a broad one which increases from 210 up to 270 K when x increases from 0.40 to 0.60, respectively. The peaks at 280, 210 and 105 K detected for the x = 0.20, 0.25 and 0.30 samples, respectively, were attributed to an AF to P transition. As was shown by Mössbauer spectroscopy these alloys are P at RT but their MS at 4.2 K (not shown here) present low magnetic fields. This low magnetic field, at this temperature, is due the AF character of the Mn atoms whose content is bigger than the Fe content in these alloys. The AF character is typical of the Fe Mn Al with high Mn content as was previously reported in melted alloys [6]. The AF character was also proved by the invariability of the MS and their HFD with and without external field at 4.2 K. The low temperature peaks detected at 50 K are attributed to a SG behavior attending the disordered character of the samples and the competitive F and AF magnetic bonds due the Fe and Mn atoms, respectively. Finally, the broad peaks detected between 210 and 270 K for the x>0.40 samples are attributed to a SP blocking behavior. This SP character was proved by Mössbauer experiments without and with an external field of 0.3 T, for the x = 0.55 sample (not shown here). It was

6 300 G. A. PÉREZ ALCÁZAR ET AL. Figure 4. Mössbauer spectra at RT of samples with different x values and after 48 hours milling. Figure 5. AC magnetic susceptibility vs. temperature for samples with different x values and 48 hours milling.

7 MAGNETIC AND STRUCTURAL PROPERTIES OF Fe Mn Al 301 Figure 6. Fraction of powder vs. particle size for the x = 0.65 sample and different preparation conditions. obtained a fitted mean field which increases from 16 to 16.8 T which is bigger than the 0.3 T value of the applied field. Also the obtained HFD of the sample with external field presents additional peaks compared with the HFD obtained without field. The extra peaks are due the extra contribution to the HF of the SP clusters which at this temperature are highly blocked and aligned to the F matrix. This behavior was also reported for melted samples with 40 at.% Al with high Fe content [6] EFFECTS OF MILLING CONDITIONS ON SAMPLES WITH LOW Al CONTENT Samples of the Fe x Mn x Al C 0.01 Cu system with 0.50 x 0.75 were milled in order to test the balls size and BM/PM effects. These alloys, when melted, present the fcc structure and the 1 at.% C and 0.3 at.% Cu addition is to improve their mechanical and corrosion properties [14]. In Figure 6 the curves of the fraction of the different powder, after sieved at different meshes, as a function of their sizes are shown, for the x = 0.65 sample. Two regions are easily observed separated by the 200 mesh size. It seems to show that this is a critical particle size on these alloys which corresponds to the most unstable size. This critical size was obtained for all preparation conditions showing that it does not depends on the balls number, on the mass, on the composition of the sample and on the milling time. It is clearly shown that for 22 hours, the higher percentage of powder corresponds to the bigger particles and this percentage decreases with the decrease of the milling time. This is expected since the MA of ductile components is a competing process between fracture and cold melting [15]. The first stages of milling are characterized by micro-forge, fracture and agglomeration which in turn produces growing of particle sizes with an irregular structure and in multilayers [16]. In Figure 7 the diffractograms corresponding to the 55, 65 and 75 at.% Fe samples are shown (those obtained for other compositions are similar). This figure illustrates the effect of the BM/PM relation and of the milling time. In Figure 7(a),

8 302 G. A. PÉREZ ALCÁZAR ET AL. Figure 7. XRD patterns of the x = 0.55, 0.65 and 0.75 samples (a) for 12 : 1 ratio and 22 hours milling, and (b) for 4 : 1 ratio and 10 hours milling. the results for samples with the 12 : 1 ratio and 22 hours of milling are shown for all compositions. They are alloyed on a fcc phase. Figure 7(b) shows, for the 4 : 1 ratio and 10 hours samples, the peaks due to the fcc structure and others due to a Al Fe alloy. Then for these conditions the powders are still under the alloying process due the low BM/PM relation and milling time. In order to explain this difference we need to remember that the mechanical alloying of elemental powders is a competitive process between fracture and cold melting of the particles during which the diffusion of the atoms is produced by the excess of point and other lattice defects generated and by momentary increase in temperature of the particles during collision. Then, higher BM/PM relations and milling times improve the alloying process. In Figure 8 diffractograms for the 4 : 1 ratio and 10 hours milling sample, in the range, are shown. This figure illustrates the effect of the composition on these samples. It is clearly observed (after a smoothing with Fourier transform and fitting with Lorentzians) the (300) peak of the α-mn, the (1 1 1) of the fcc alloy and the (1 1 0) peak attributed to the Al Fe alloy poor in Al (we have low Al content). Furthermore, it is shown that with the increasing of Fe content,

9 MAGNETIC AND STRUCTURAL PROPERTIES OF Fe Mn Al 303 Figure 8. XRD patterns for samples with different x values, a 4 : 1 ratio and 10 hours milling. the Mn peak decreases until disappear and the Al Fe one raises. Also, it is shown that the widespread phase for low Fe contents is the fcc one, and that for high Fe contents this phase competes with the bcc one of the Al Fe alloy. Finally, and in order to illustrate the ball size effect in the milling products, Mössbauer spectra were taken in samples milled during 10 hours using balls of 1 and 2 cm diameter, respectively, and BM/PM of 4 : 1 and 12 : 1. Figure 9 shows the Mössbauer spectra of these samples. Figure 9(a) shows the spectra corresponding to samples with 50 at.% Fe. It can be noted that the spectrum for the 4 : 1 ratio presents two contributions, a sextet and a broad single line. The best fit was obtained with two hyperfine field distributions, one with big HFs and other with low HFs. The spectral area of the HFD with big fields is bigger when big balls are used and is lower for small balls. The spectrum for the 12 : 1 relation and big balls was fitted with two HFDs (with big and low fields, respectively) and that for 12 : 1

10 304 G. A. PÉREZ ALCÁZAR ET AL. Figure 9. Mössbauer spectra for two different ratios and ball sizes for (a) x = 0.50 and (b) x = relation and small balls with a HFD of low fields. The obtained XRD patters (not showed here) showed that the first three samples present the ternary fcc phase and a bcc phase corresponding to a Fe Al alloy poor in Al and that the last one presents only the fcc phase. Then, it can be attributed the HFD with high fields to the Fe Al phase which is F [13] and the HFD with low fields to the fcc phase which is AF [7]. These results show, remembering that the stable phase for this composition in melted alloys is the fcc, that the alloying process is more efficient for small balls and relation 12 : 1. Figure 9(b) shows the spectra corresponding to samples with 75 at.% Fe. In this case the first three spectra were fitted with a HFD and a single paramagnetic line and the last one with a single paramagnetic line. The obtained XRD results (not shown here) showed that the three first samples present the fcc phase and the Fe Al alloy and that the last one has only the fcc ternary alloy. In this way, samples with 75 at.% Fe in the bcc phase continues to be F but the ternary fcc phase is P. This paramagnetic behavior of the fcc phase is attributed to the low Mn content which is the atom which induces the AF behavior in this structure [7]. Acknowledgements Authors would like to thank Colciencias (Colombian Agency) and Universidad del Valle for financial support.

11 MAGNETIC AND STRUCTURAL PROPERTIES OF Fe Mn Al 305 References 1. Chakrabarti, D. J., Metall. Trans. B 8 (1977), Kobeissi, M. A., J. Phys.: Condens. Matter 3 (1991), Bremers, H., Jarms, Ch., Hesse, J., Chajwasiliou, S., Efthiadis, K. G. and Tsonkalas, Y., J. Magn. Magn. Matter 140 (1995), Zamora, L. E., Pérez Alcázar, G. A., Bohórquez, A., Marco, J. F. and González, J. M., J. Appl. Phys. 82 (1997), Zamora, L. E., Pérez Alcázar, G. A., Tabares, J. A., Bohórquez, A., Marco, J. F. and González, J. M., J. Phys. Condens. Matter 12 (2000), González, C., Pérez Alcázar, G. A., Zamora, L. E., Tabares, J. A. and Greneche, J. M., J. Phys. C 14 (2002), Pérez Alcázar, G. A., Galvão da Silva, E. and Paduani, C., Hyp. Interact. 66 (1992), Sánchez, A. J., Bohórquez, A. and Pérez Alcázar, G. A., Hyp. Interact. (C) 2 (1997), Ishikawa, Y. and Endoh, Y., J. Phys. Soc. Jpn. 23 (1967), Medina, G., Pérez Alcázar, G. A., Muriel, A. and Tabares, J. A., Hyp. Interact. (C) 4 (1999), Cullity, B. D., Elements of X-Ray Diffraction, Adisson-Wesley Pub. Co., Rico, M. M., Medina, M. H. and Pérez Alcázar, G. A., Hyp. Interact. 128 (2000), Vincze, I., Phys. Status Solidi A 7 (1971), K Rodríguez, V. F., Jiménez, J. A., Adeva, P., Bohórquez, A., Pérez, G. A., Fernández, B. J. and Chao, J., Rev. Metal. Madrid 34 (1998), Benjamin, J. S., Sci. American 5 (1976), Cruz, B., Pérez Alcázar, G. A. and Aguilar, Y., Phys. Stat. Solidi (b) 220 (2000), 381.