English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
Introduction
Many research groups in the entire world are currently working on the development of new alloys with soft magnetic properties for technological applications in spintronic, biomedicine, magnetic sensors, and energy. The full Heusler alloys or compositions close to them, in particular, have systematically attracted the interest of the scientific community due to their great variety of magneto-structural effects, such as: high magnetocaloric, ferromagnetically shaped memory, semi-metallicity, exchange bias, magneto-resistive, among others [1-12].
In general, Heusler alloys of the X2YZ type (full-Heusler alloy) consist of two transition metal atoms, X and Y, with the number of electrons d : n d > 5 and n d ≤ 5, respectively, and one non-metal atom, Z, with the number of d electrons nd = 0 or 10, and hybridizations of the sp-type. At room temperature (RT), the X2YZ Heusler alloys have a centered body cubic structure, L21 -type (Fm3m), with the Y and Z atoms interconnected between two tetrahedral lattices, occupying the centers of each cube of X atoms (austenitic phase).
In the case of Heusler ferromagnetic alloys with Y = Mn [13-16], the Mn atoms may usually play the role of electron spin-polarizer. Otherwise, the X atom is responsible for changing the lattice parameter of the L21 phase, as well as providing the localized electrons (3d) and conduction electrons (4s), responsible for the localized and itinerant magnetism, respectively. Z atoms act as non-magnetic spacers regulating the exchange interaction, constituting a source of sp electrons that participate in the magnetic interaction between Mn atoms (indirect exchange coupling). In the case of Mn-based Heusler alloys, their magnetic properties depend primarily on the distance between manganese atoms (Mn-Mn), which in turn, depends on the lattice parameter of the L21 phase.
For instance, studies on Fe2MnSn Heusler alloys with compositions close to the full Heusler have recently been reported [17-19]. For example, V. K. Jain et al. [17] have studied the nanocrystallization - by mechanical grinding (high energy ball-milling)- of the Fe2MnSn Heusler alloy, previously prepared by melting a blend of elemental powders. It has combined experimental (analysis of X-ray diffraction and DC magnetization measurements) and also theoretical (calculations of state density and band structure) approach to show that the initial molten sample has presented an L21-type majority phase (full Heusler structure), with a lattice parameter equal to 6.246 Å, co-existing with a tetragonal-like phase. However, the nanocrystallization process of the samples has shown a gradual reduction of the tetragonal phase, but there is a gap in the literature related to the investigation of order-disorder phases as well as magnetic properties of Fe2MnSn alloys in different crystalline structures.
In the present work, we studied the micro-granular structure of the arc-melted and annealed Fe50Mn25+xSn25-x alloys with x = -1.25, 0.0, 2.5, 5.0, 7.5 (close to the full-Heusler alloys). It has been shown that the L21-type structure is not stabilized. However, due to rapid cooling (quenching), the formation of the ε-(Fe/Mn)3Sn and ε-(Fe/Mn)3(Sn/Fe/Mn) phases are favored, both with hexagonal D019-type symmetry (space group: P63/ mmc), where the Mn atoms randomly share the sites (6h) with Fe and Sn. Local (57Fe and 119Sn Mössbauer spectroscopies) and bulk (X-ray diffraction and magnetization) measurements were employed to understand the structural and magnetic properties of these alloys.
1. Experimental procedure
1.1. Sample preparation
The Fe50Mn25+xSn25-x system, with x -1.25, 0.0, 2.5, 5.0, 7.5 (the minus sign "-" corresponds to a fraction of Mn that is subtracted from Heusler's composition) was prepared using an arc-melting furnace under a controlled and high purity argon atmosphere. Five Fe-Mn-Sn alloys were prepared from mixtures of high purity elemental powders of Fe (99.5%), Mn(99.9%), and Sn(99.9%). In order to ensure homogeneity of the alloys, the melting process of each mixture was repeated, at least, five times, rotating the solidified sample inside the casting chamber. The mass of each prepared solid sample was approximately 2 g (the loss of mass after the melting process is less than 2 %). Subsequently, each solid sample (previously sealed in a quartz tube, under argon atmosphere) was annealed at 1173 K for 4 days and quenched in iced-water (a mixture of water and ice). The sample preparation process followed a methodology previously developed in our group, which guarantees the reproducibility of sample properties.
1.2. X-ray diffraction
The structural properties of the annealed alloys were analyzed by X-ray diffraction technique using a Bruker-D8 Focus diffractometer, operated at 40 kV and 40 mA with Cu-Kα radiation ((λ= 1.5406 Å). The percentage of the phases for each alloy, as well as the microstructural parameters (lattice parameters, atomic occupations, phase fraction), were obtained by Rietveld refinement using the Topas software (Bruker AXS) [14]. The crystallite sizes and their internal micro-tensions were separately calculated by the Williamson-Hall method.
1.3. Mössbauer spectroscopy
The local magnetic orders of the annealed Fe50Mn25+xSn25-x alloys were systematically studied by 57Fe and 119Sn Mössbauer spectroscopies. The 57Fe and 119Sn isotopes act as local probes in the study of the localized magnetism from the d-electrons of the Fe atoms and the itinerant magnetism of the conduction s-electrons, respectively. Likewise, considering that the spin polarization process is a global-like effect, Fe and Mn atoms also contribute to the itinerant magnetism character measured in the Sn sites (119Sn isotopes only probe the transferred magnetic hyperfine fields).
The Fe and Sn Mössbauer spectra were obtained for all samples at RT using a transmission Mössbauer spectrometer, with standard radioactive sources of 57Co/Rh (25 mCi) and 119mSn/CaSnO3 (5 mCi), respectively. The Mössbauer parameters [isomer shift (δ), magnetic hyperfine field (B h f ), linewidth (Γ), and absorption area (A)] in both Fe and Sn sites were obtained using the NORMOS program. The δ -values are relative to radioactive sources.
1.4. Magnetization measurement
Magnetic hysteresis loops [M(H)] were recorded at RT for all annealed Fe50Mn25+xSn25-x alloys using a vibrant sample magnetometer (VSM) from the Faculty of Sciences of the National University of Engineering. This is a local manufacturing equipment, with automatic measurement control, characterized by a sensitivity of 10-4, capable of measuring magnetic moments on the order of 10 emu, under a maximum magnetic field of 2300 Oe. Calibration was performed using a Ni (99.9 %) disc (38 mm diameter, 120 mm thick, and 30.7 mg mass).
2. Results and discussion
2.1. Rietveld analysis of the XRD patterns of the Fe50Mn25+xSn25-x powder alloys with x: -1.25, 0.0, 2.5, 5.0, 7.5
X-ray diffraction (XRD) patterns of the annealed Fe50Mn23.75Sn26.25 Fe50Mn25Sn25, Fe50Mn27.5Sn22.5, Fe50Mn30Sn20, and Fe50Mn32.5Sn17.5 alloys are respectively shown in Figures 1 (a), (b), (c), (d), and (e). The Bragg peaks of the crystallographic phases, identified by Rietveld refinement, are also indicated in the XRD patterns.
FIGURE 1 X-ray diffraction patterns of the Fe50Mn25+xSn25-x alloys with: (a) x = -1.25, (b) x = 0, (c) x= 2.5, (d) x = 5, (♦) x = 7.5. (+) [ε - (Fe/Mn) 3 Sn], (Δ) [ε-Fe 5 Sn 3 ], and (•) [ε - (Fe/Mn) 3 (Sn/Fe/Mn)].
Table 1 shows the structural parameters [lattice constant, average crystallite size <D>, internal strain (< ε2 >1/2), atomic occupancy (Ocp), and phase fraction (Xph )] obtained for each alloy from the Rietveld refinement and the Williamson-Hall method. The quality of Rietveld refinements is quantified by the low values of chi-squared (x 2)[x 2 = (R wp /R exp ) 2 , where R wp and R exp are weighted profile R-factor and expected R-factor factors, respectively]. The different occupations of the shared sites reflect the atomic disorder in each alloy.
TABLE 1 Structural parameters obtained from Rietveld refinement of the XRD patterns of the annealed Fe50Mn25+xSn25-x alloys with: (a) x = -1.25, (b) x = 0, (c) x = 2;5, (d) x = 5, (e) x = 7.5 [lattice constants (a, b, c), crystalline size <D>, internal strain (< ε2 > 1/2 ), atomic occupancy (Ocp), and percentage of the phases (Xph)].
The XRD patterns of the Fe50Mn23.75Sn26.25 and Fe50Mn25Sn25 alloys were fitted using two hexagonal crystalline phases: (i) a crystalline solid solution identified as the ε-(Fe/Mn)3Sn (majority phase), and (ii) the ε-Fe5Sn3 intermetallic compound (minority phase). The ε -(Fe/Mn)3Sn solid solution has a hexagonal structure D019 (spatial group P63/mmc), where Fe and Mn atoms share 6h Wyckoff sites and Sn atoms are located at 2c site.
The ε-Fe5 Sn3 intermetallic is described by the P6 3 /mmc spatial group, where Fe atoms are placed in two 2a and 2d Wyckoff sites. Moreover, Sn atoms are located at the 2c site. Figures 2 (a) and (b) show the observed (black) and calculated (red curve) XRD patterns of the Fe50Mn23.75Sn26.25 and Fe50Mn25Sn25 alloys, respectively. Below each XRD pattern, the difference in the intensities between observed and calculated diffraction patterns is also plotted.
FIGURE 2 Rietveld refinement of the X-ray diffraction patterns of the annealed (a) Fe50Mn23.75Sn26.25 and (b) Fe 50 Mn 25 Sn 25 alloys. Positions of the Bragg-reflections are shown for the (♦) ε-(Fe/Mn) 3 Sn, and (Δ) ε -Fe 5 Sn 3 phases. The difference between observed and calculated patterns and the calculated reflection positions (green color on bottom trace) is also displayed.
The first conclusion is that the Fe50Mn25Sn25 alloy, despite X2YZ composition, does not correspond to a full-Heusler alloy characterized by an L21-type structure. The XRD patterns of the Fe50Mn27.5Sn22.5 , Fe50Mn30Sn20 , and Fe50Mn32.5Sn17.5 alloys were fitted using a ε-(Fe/Mn)3(Sn/Fe/Mn) hexagonal phase. It is described by P63/mmc space group, where Fe and Mn atoms randomly share the 6h sites. Furthermore, other atoms of Fe, Mn, and Sn share 2c sites. The difference between the observed and calculated XRD patterns for these alloys is also shown at the bottom of each XRD pattern displayed in Figures 3.
FIGURE 3 Rietveld refinement of the X-ray diffraction patterns of the annealed (a) Fe 50 Mn 27.5 Sn 22.5 , (b) Fe50Mn30Sn20, and (c) Fe50Mn32.5Sn17.5. Positions of the Bragg-reflections are shown for the ( • ) ε - (Fe/Mn) 3 (Sn/Fe/Mn) phase. The difference between observed and calculated patterns and the calculated reflection positions (green color on bottom trace) is also displayed.
The average crystallite sizes, estimated from the Williamson-Hall method, are in the range of 256-62 nm, indicating that some of our samples are in the nanoscale regime (< 100 nm), while others are in micrometer size regime (bulk-like phase). A reduction in average sizes was observed as the Mn concentration increases due to a reduction in the number of nuclei during fast cooling process (see Table 1).
2.2. 57Fe Mössbauer study and magnetic hyperfine properties
RT 57Fe Mössbauer spectra of the annealed Fe50Mn23.75Sn26.25, Fe50Mn25Sn25 , Fe50Mn27.5Sn22.5 , Fe50Mn30Sn20 , and Fe50Mn32.5Sn17.5 alloys are respectively plotted in the Figures 4 (a), (b), (c), and (d). These spectra were fitted assuming the crystalline phases obtained from the X-ray analyses.
FIGURE 4 RT 57 Fe Mössbauer spectra of the annealed Fe50Mn25+xSn25-x alloys with: (a) x = -1.25, (b) x = 0, (c) x = 2;5, (d) x = 5, (e) x = 7.5. Magnetic hyperfine field distribution P(B hf ) curves, obtained from the fits of the corresponding Mössbauer spectra, are plotted on the right-hand side for each one.
The 57Fe Mössbauer spectra, in general, display broad six-line patterns due to nuclear Zeeman interaction [an interaction between nuclear magnetic moment and magnetic hyperfine field (Bhf) at Fe nucleus produced by s polarized electrons and 3d unpaired electrons]. The broadening line shapes of the 57Fe Mössbauer spectra indicate that the Fe atoms are magnetically ordered, but there is a distribution of Fe non-equivalent sites. This sub-spectra were fitted using a distribution of 40 magnetic hyperfine fields P(Bhf). To account for small asymmetries of the spectra, a linear correlation between Bhf and the isomer shift (δ) was assumed for the 40 magnetic sub-spectra.
The 57Fe Mössbauer spectra of the annealed Fe50Mn23 75Sn2625 and Fe50Mn25Sn25 alloys in Figures 4 (a) and (b) were fitted using a distribution of magnetic hyperfine magnetic fields P(Bhf) with 40 sub-spectra, which was associated to phase ε-(Fe/Mn)3Sn. The fractions of the crystalline sites of the Fe5Sn3 phase at 20.0 T, 20.8 T, and 24 T are less than 7%.
The 57Fe Mössbauer spectra of the annealed Fe50Mn27.5Sn22.5, Fe50Mn30Sn20, and Fe50Mn32.5Sn17.5 alloys were also fitted using a magnetic hyperfine distribution P(Bhf) associated with the Fe magnetic sites in the ε- (Fe/Mn)3(Sn/Fe/Mn) hexagonal phase and a broad singlet. Indeed, the singlet must be associated with those Fe atoms in a paramagnetic state, which share the 2c sites with the Sn and Mn atoms. The fraction of the singlet component increases as the x-content rises. Table 2 shows the average values of the magnetic hyperfine field (Bhf and B tr hf) and the isomer shift (δ Fe and δ Sn ) of non-equivalent Fe and Sn sites.
TABLE 2 Summary of magnetic parameters determined at RT for the annealed Fe50Mn25+xSn25-x con: (a) x = -1.25, (b) x = 0, (c) x = 2;5, (d) x = 5, (e) x = 7.5 (magnetic hyperfine field (B hf ), maximum magnetization (M s ), coercive field (H c ), remnant magnetization (M r ) and exchange bias filed (Heb)). δ Fe and δ Sn are the isomer shift values in Fe-site and Sn-site for each alloys.
2.3. Mössbauer study of the transferred magnetic fields in 119Sn sites
Figures 5 (a), (b), (c), (d), and (e) display the RT 119Sn Mössbauer spectra for the annealed Fe50Mn23.75Sn26.25, Fe50Mn25Sn25, Fe50Mn27.5Sn22.5, Fe50Mn30Sn20, and Fe50Mn32.5Sn17.5 alloys, respectively. These 119Sn Mössbauer spectra also show a distribution of magnetic hyperfine fields with absorption peaks between -10 mm/s and +12 mm/s. So, it clearly indicates the presence of transferred hyperfine magnetic fields (Bhf) at the Sn sites from the conduction electron polarization effect caused by Fe and Mn atoms nearby.
FIGURE 5 RT 119 Sn Mössbauer spectra of the annealed Fe50Mn25+xSn25-x with: (a) x = -1.25, (b) x = 0, (c) x = 2;5, (d) x = 5, (e) x = 7.5. Magnetic hyperfine field distribution P(B tr hf ) curves, obtained from the fits of the corresponding Mössbauer spectra, are plotted on the right-hand side for each one. The B tr hf is the transferred hyperfine magnetic fields from polarized conduction s-electrons by Fe and Mn atoms. The red arrows indicate the position of paramagnetic Sn component.
The spectra were also fitted with a distribution of magnetic hyperfine fields, but now due to the transferred magnetic fields at Sn sites. A magnetic ordering of the Fe and Mn magnetic moments was assumed and then via RKKY interactions, the s-type conduction electrons were polarized and have lead the information to the Sn sites.
In Figures 5 (a) and (b), the 119Sn Mössbauer spectra of the annealed Fe50Mn23.75Sn26.25 and Fe50Mn25Sn25 alloys were respectively fitted using a P(B tr hf ) distribution with 40 magnetic sub-spectra. Each sub-spectrum corresponds to the transferred magnetic fields B tr hf at Sn sites due mainly to the magnetic moments of the neighboring Fe and Mn atoms of phase e-(Fe/Mn)3Sn. The 119Sn Mssbauer spectra of the annealed Fe50Mn27.5Sn22.5, Fe50Mn30Sn20, and Fe50Mn32.5Sn17.5 alloys are shown in Figures 5 (c), (d), and (e). These spectra were analyzed using a P(Bhf) distribution of 40 magnetic sub-spectra associated with B tr hf fields at Sn sites due to s-type electrons polarized by Fe and Mn atoms in the hexagonal phase ε-(Fe/Mn)3(Sn/Fe/Mn). In addition, in the central part of the spectra corresponding to the samples with x = 5.0 and 7.5, it is possible to observe a new contribution at a velocity of 1.8 mm/s (see arrows in Figures 5(d) and (e)), indicating that a small fraction of Sn atoms fees zero B tr hf value. Thus, it can be noticed that there are regions with only Sn atoms or regions where the transferred fields of Fe and Mn are compensated (a singlet was also observed in 57Fe sites).
2.4. RT magnetization measurements
Figure 6 shows the M(H) loops of the annealed Fe50Mn25+xSn25-x alloys. The inset presents an amplification of the central part where a horizontal shift of the hysteresis curves relative to the zero-field was observed. The observed M(H) loop shift effect (not fully understood yet in our samples) may be associated with the exchange bias effect due to the incipient formation of an antiferromagnetic phase, that arises as the Mn-Mn distances changes and by Mn excess [20]. Table 2 shows a comparison of the magnetic parameters of each alloy obtained from magnetization measurements and from 57Fe and 119Sn Mössbauer spectroscopies.
FIGURE 6 Hysteresis curves of magnetization (M) vs. applied field (H) recorded at RT of the annealed Fe 50 Mn 25 +xSn 25-x alloys with: (a) x = -1.25, (b) x = 0, (c) x = 2;5, (d) x = 5, (e) x = 7.5. The inset shows the central area of the hysteresis curves of all alloys in order to highlight the horizontal shift of the curves and also the remnant and coercive fields.
It is observed that all annealed alloys exhibit a soft ferromagnetic behavior, characterized by low values of the coercive (9.8-50.4 Oe) and remnants fields (2-7 emu/g). On the one hand, the maximum magnetization Ms (obtained with 2.2 kOe field) gradually decreases from 204.3 to 106.9 emu/g as the x-fraction of Mn increases. On the other hand, the coercive field Hc increases with the value of x, showing that the magnetic properties depend on the substitution process of Sn atoms by Mn. The highest increase of the coercive field values in the alloys with x = 5.0 and 7.5 is associated with the microstructural deformation and defect generation by reduction of crystallite sizes, as shown in the XRD data. In a first approximation, there is a relationship between hyperfine magnetic field Bhf and Fe magnetic moments, which allows establishing some comparisons with magnetization measurements. From this comparison, it is observed that the results of the magnetization measurements are in agreement with the results from the 57Fe Mössbauer spectroscopy that show a gradual increase of the singlet fraction as the x-content is increased (indirectly the Sn paramagnetic component also increases). In other words, it is also possible to establish that the observed singlets in the 119Sn and 57Fe Mössbauer spectra of the annealed samples with x = 2.5, 5.0, 7.5 are responsible for diminishing the saturation magnetization Ms when x-fraction increases.
Conclusions
Combining results of X-ray diffraction, 119Sn and 57Fe Mössbauer spectroscopies, and magnetization, it has been determined the structural and magnetic properties of the arc-melted and annealed Fe50Mn25+xSn25_x Heusler alloys with x = -1.25, 0.0, 2.5, 5.0, 7.5. These compositions are close to those of full-Heusler Fe2MnSn alloy, but the L21-like cubic structure (the fingerprint of the full-Heusler) was not stabilized due to the fast quenching process (it is assumed that the thermal annealing at 1173 K and subsequently rapid cooling in iced water is responsible for the absence of the L21-like cubic structure). The Rietveld refinement analyses of all diffractograms of the annealed Fe50Mn25+xSn25_x alloys have shown the formation of hexagonal phases D019 (spatial group P6 3 /mmc). For annealed samples with x = -1.25 and 0.0, the XRD patterns have indicated two crystalline phases: (i) a disordered solid solution e-(Fe/Mn)3Sn (majority) and (ii) the intermetallic compound ε-(Fe)5Sn3, whereas the alloys with x = 2.5, 5.0 and 7.5 have only one hexagonal phase, corresponding to a disordered crystalline solid solution, e- (Fe/Mn)3(Sn/Fe/Mn). The average crystallite sizes (calculated from Williamson-Hall method) are within the range of 62-256 nm, showing that some of our alloys are in a nanoscale regime, while others are in a bulk-like one (micrometer regime). The magnetic properties of these materials are typically similar to those found for soft magnets. They have shown low coercive and remnant fields. An increase in the coercive field of the annealed Fe50Mn25+xSn25_x alloys is observed when the Mn fraction is increased in opposite to the behavior found for the maximum magnetization recorded at 2200 Oe. The reduction of the maximum magnetization (recorded at 2200 Oe) by increasing the atomic fraction of Mn is accompanied by the increase in the non-magnetic fraction of the 57Fe and 119Sn Mössbauer spectra, more pronounced in 57Fe sites.
Furthermore, the Mössbauer results of the 57Fe and 119Sn isotopes of all annealed alloys have shown the following characteristics: (i) localized magnetism in chemically disordered systems from distribution of magnetic hyperfine fields at the Fe sites, and (ii) of itinerant magnetism originated by transferred hyperfine fields from the neighboring Fe and Mn atoms in the 119Sn nuclei. Alloys with x = 2.5, 5.0, and 7.5 have also shown a small non-magnetic fraction due to regions with a higher concentration of Sn or regions where the transferred field of the Fe and Mn atoms is canceled.
