2021 Volume 62 Issue 6 Pages 887-891
Sm–Fe–M (M = Ti, V) sintered magnets with the ThMn12-type structure which consist of higher Sm content than the stoichiometric composition of ThMn12-type compound, were produced by the powder-metallurgy. Sm–Fe–Ti sintered magnets had a low level of coercivities of about 10 kAm−1. In contrast, Sm–Fe–V anisotropic sintered permanent magnets, particularly, Sm10.5Fe74.1V15.4 magnet exhibited magnetic properties with a coercivity of 648.7 kAm−1, a remanence of 0.73 T and a (BH)max of 97.6 kJm−3. Through microstructure observation of the grain boundary between the adjacent ThMn12-type grains using STEM, no grain boundary phases were observed in Sm10.0Fe81.7Ti8.3 magnet, whereas two types were observed in Sm10.5Fe74.1V15.4 magnet: an amorphous structure with a thickness of about 1 nm and an incomplete periodic structure with a thickness of about 2 nm. EDS analysis indicated that the composition ratios of Sm to the sum of Fe and V were about 1 to 1 and 1 to 1.7, respectively. It suggested that the existence of a grain boundary phase is important to raise coercivity.
Fig. 3 Demagnetization curves of Sm–Fe–V sintered magnets.
The Nd–Fe–B magnet has played a very important role as a key material for electric and electronic industries and is forecasted to expand the demand due to promoted use of electric vehicles by economic policies for environmental issues. However, to focus on the field of resources, an increase in the demand of the Nd–Fe–B magnet would lead to an increase in the usage of rare-earth elements. Thus, it is assumed that there will be an unstable supply not only of heavy rare-earth elements but also of light rare-earth elements, namely Nd, and it is important to propose the use of a variety of permanent magnets to reduce the quantity of some rare-earth elements being used.
ThMn12-type compounds discovered by Ohashi et al. in the late 1980s, is refocused as hard magnetic materials with reduced rare-earth content. The SmFe11Ti compound showed a higher anisotropy field but a lower saturation magnetization than those of the Nd2Fe14B compound.1) Recently, Hirayama et al. achieved to produce Sm(Fe0.8Co0.2)12 thin films with intrinsic magnetic properties exceeding those of the Nd2Fe14B compound, but no coercivities were shown.2) Ogawa et al. reported that the Sm(Fe0.8Co0.2)12 anisotropic polycrystalline thin film showed a coercivity of 382 kAm−1 which increased up to 668 kAm−1 after Cu–Ga diffusion process.3) As a new report by Sepehri-Amin et al., B doped Sm(Fe0.8Co0.2)12 anisotropic polycrystalline thin film in which B is segregated at the grain boundary showed a high coercivity of 955 kAm−1 with a high remanence of 1.5 T.4) These reports suggest a strong relationship of the coercivity and the grain boundary phases between the ThMn12-type grains. However, the Sm(Fe0.8Co0.2)12 compound has unfortunately not been realized to form the bulk magnet because of being thermodynamically unstable.
Some attempts have been made to produce a bulk magnet for the ThMn12-type compound with stabilizing elements such as Ti, V, Mo, Cr, W and Si so far. Ohashi, Hadjipanayis et al. and Wang et al. reported on Sm–Fe–Ti or Sm–Fe–Ti–V sintered magnets, but these showed few coercivities.5–7) Sugimoto et al. reported on Sm–Fe–V homogenized ingot with a coercivity of 294 kAm−1.8) Pinkerton and Wingarden, Schultz et al. and Schönhöbel et al. reported on Sm–Fe–V bulk magnets made by hot-pressing, resin-bonding, hot-compacting and then hot-deforming process with high coercivities 446–891 kAm−1, whereas remanences of them with under 0.5 T were much lower than saturation magnetic polarization because of isotropic magnets.9–11)
The purpose of this study is to prepare the anisotropic sintered permanent magnets with the ThMn12-type compound. As well as the microstructural features in the Nd–Fe–B sintered magnets, Sm(Fe, M)12 based sintered magnets are supposed to have minor phases which surround the Sm(Fe, M)12 grains. Therefore, the phase equilibrium of Sm-rich off-stoichiometric alloys is very important which will change depending on the stabilizing element M.
The sintered magnets were obtained by the powder-metallurgy. Sm, Fe, Ti metals and ferrovanadium were weighed to give the initial composition with excess Sm to compensate for the loss of Sm during melting and to form Sm-rich minor phase. These raw materials were melted by induction melting and were casted as a starting alloy. Unlike the Sm–Fe–Ti as-cast ingots, the Sm–Fe–V ingots contains primary α-(Fe, V) and were homogenized at 900°C for 40 h in argon atmosphere. The coarse powder obtained from the ingot by hydrogen decrepitation was mixed with a lubricant and jet-milled to an average particle diameter of a few microns in nitrogen. The fine powder thus obtained was then oriented and compacted in a magnetic field in a nitrogen atmosphere. The compacts were sintered at 1140 to 1200°C for 3 h, and then were quenched in an argon chamber, by using a furnace with two chambers.
Chemical compositions of sintered magnets were quantified by ICP-AES. The crystal structures were identified by X-ray diffraction (XRD) with Cu-Kα radiation. Sintered magnets were machined into predetermine dimensions and were magnetized in a pulsed magnetic field of 3.8 MAm−1. The magnetic properties were measured using a permeameter and a pulse BH loop tracer with maximum magnetic field of 1.6 MAm−1 and 6.0 MAm−1, respectively. The microstructures and chemical compositions of the phases were investigated using an FE-EPMA, an EBSD and an HAADF-STEM with EDS. Thin specimens for the STEM observation were prepared by an FIB-SEM.
Table 1 shows sintering temperatures, densities and chemical compositions of Sm–Fe–M sintered magnets with various Sm contents. Sm–Fe–Ti magnets are sintered at 1180°C or 1200°C and Sm–Fe–V magnets are sintered at 1140°C to densify the sintered bodies. Sm contents were varied from 8.6 to 10.0 at% for Sm–Fe–Ti magnets and from 8.6 to 10.5 at% for Sm–Fe–V magnets, while Ti and V contents were constant at approximately 8 at% and 15 at%, respectively. The density increases with the Sm content increases for Sm–Fe–Ti and Sm–Fe–V. Figure 1 shows the XRD patterns taken from powdered Sm10.0Fe81.7Ti8.3 and Sm10.5Fe74.1V15.4 sintered magnets. The main phase of these samples can be indexed as the ThMn12-type structure with a small amount of Sm oxycarbide of which O and C are the typical impurities during the powder-metallurgy process with a lubricant.
XRD patterns of powdered Sm10.0Fe81.7Ti8.3 and Sm10.5Fe74.1V15.4 sintered magnets.
Figure 2 shows demagnetization curves of Sm–Fe–Ti sintered magnets with various Sm contents measured along the parallel direction to the alignment field. Remanences are 0.2 to 0.4 T which decreased with increasing Sm content, and coercivities are about 10 kAm−1 regardless of Sm content. Sm–Fe–Ti sintered magnets have poor magnetic properties compared with the anisotropy field (8.21 MAm−1) and the saturation magnetic polarization (1.26 T) of the SmFe11Ti compound reported by Kuno et al.12)
Demagnetization curves of Sm–Fe–Ti sintered magnets.
Figure 3 shows demagnetization curves of Sm–Fe–V sintered magnets with various Sm contents measured along the parallel direction to the alignment field. Sm8.6Fe76.5V14.9 and Sm9.6Fe75.6V14.8 with low Sm content have low coercivities of 44.8 kAm−1 and 58.3 kAm−1 and low remanences of 0.63 T and 0.62 T, respectively. In contrast, Sm9.9Fe75.2V14.9 and Sm10.5Fe74.1V15.4 with high Sm content exhibit higher coercivities of 670.3 kAm−1 and 648.7 kAm−1 and remanences of 0.60 T and 0.73 T, respectively. The (BH)max reaches 97.6 kJm−3 with Sm10.5Fe74.1V15.4 magnet. These magnetic properties are one of the largest values reported so far as the ThMn12-type sintered magnets.
Demagnetization curves of Sm–Fe–V sintered magnets.
Figure 4 shows hysteresis loops of the Sm10.5Fe74.1V15.4 magnet measured along the parallel and perpendicular directions to the alignment field, measured using a pulse BH tracer. The anisotropic feature of the magnet is due to the alignment of the ThMn12-type particles during the pressing process in a magnetic field. The magnetic polarization at the applied field of 6.0 MAm−1 is 0.82 T and the anisotropy field is expected over 6.0 MAm−1.
Hysteresis loops of the Sm10.5Fe74.1V15.4 magnet measured along parallel and perpendicular directions to the alignment field.
Figure 5 shows BSE images of Sm–Fe–Ti sintered magnets observed by FE-EPMA. The gray contrast is the ThMn12-type main phase, whose average grain size is 7.5 µm estimated from the EBSD image (not shown in this paper) of the Sm9.7Fe82.2Ti8.1 magnet. The white contrast shows a Sm oxycarbide phase. In the Sm8.6Fe83.3Ti8.2 magnet (Fig. 5(a)), the α-Fe phase containing Ti (hereinafter referred to as α-(Fe, Ti)) with a size of a few microns (the dark gray contrast) and the TiC phase with a size of submicron (the black contrast) are found. In contrast, the formation of α-(Fe, Ti) phase, which is suggested to be due to the lack of Sm, was not found in the Sm9.7Fe82.2Ti8.1 magnet (Fig. 5(b)). Sm(Fe, Ti)2 phase containing a small amount of Ti with the shaded white contrast and a Sm3(Fe, Ti)29 phase with the bright gray contrast can be seen at the triple junction in Sm10.0Fe81.7Ti8.3 (Fig. 5(c)).
BSE images of (a) Sm8.6Fe83.3Ti8.2, (b) Sm9.7Fe82.2Ti8.1 and (c) Sm10.0Fe81.7Ti8.3 magnets.
Figure 6 shows BSE images of Sm–Fe–V sintered magnets observed by FE-EPMA. The ThMn12-type phase can be observed as the main phase and the average grain size is 9.5 µm for the Sm10.5Fe74.1V15.4 magnet. The Sm oxycarbide phase are also observed in the Sm–Fe–V magnets. In the Sm8.6Fe76.5V14.9 magnet (Fig. 6(a)), the α-Fe phase containing V (hereinafter called α-(Fe, V)) having a few microns with the black contrast is found. As observed in the Sm–Fe–Ti magnets, the Sm9.9Fe75.2V14.9 (Fig. 6(b)) and the Sm10.5Fe74.1V15.4 (Fig. 6(c)) magnets with higher Sm content are free of the α-(Fe, V) phase. In the Sm10.5Fe74.1V15.4 (Fig. 6(c)), a Sm(Fe, V)2 phase containing a small amount of V with the light gray contrast and a Sm-rich phase with the bright white contrast can be observed at the triple junction. The Sm-rich phase is not found in the Sm–Fe–Ti magnets.
BSE images of (a) Sm8.6Fe76.5V14.9, (b) Sm9.9Fe75.2V14.9 and (c) Sm10.5Fe74.1V15.4 magnets.
Figure 7(a) shows an HAADF-STEM image of an FIB fabricated thin specimen of the Sm10.0Fe81.7Ti8.3 sintered magnet. The image of the grain boundary shows that two ThMn12-type grains contact directly without any grain boundary phase. Figure 7(b) shows EDS line profiles for Fe, Sm and Ti concentrations from selected 20 points indicated by white line in Fig. 7(a). Sm content is not changed at the grain boundary. It can be said that the grain boundary phase between the ThMn12-type grains is none or unobservable in the Sm10.0Fe81.7Ti8.3 magnet although it contains Sm-rich compounds such as Sm(Fe, Ti)2 and Sm3(Fe, Ti)29.
(a) An HAADF-STEM image of an FIB fabricated thin film of the Sm10.0Fe81.7Ti8.3 sintered magnet, (b) EDS line profiles for Fe, Sm and Ti concentrations from selected 20 points in (a).
Figure 8(a) and (c) show HAADF-STEM images of an FIB fabricated thin specimen of the Sm10.5Fe74.1V15.4 sintered magnet. Grain boundary phases between the ThMn12-type grains can be clearly observed. In Fig. 8(a), the grain boundary phase with a thickness of approximately 1 nm can be said an amorphous structure. Figure 8(b) shows EDS line profiles for Fe, Sm and V concentrations in Fig. 8(a). Sm is enriched at the grain boundary phase, up to 50.6 at% while Fe and V are lean. The composition ratio of Sm and (Fe + V) is approximately 1:1.
(a) and (c) HAADF-STEM images of an FIB fabricated thin film of the Sm10.5Fe74.1V15.4 sintered magnet, (b) and (d) EDS line profiles for Fe, Sm and V concentrations from selected 20 points in (a) and (c).
In Fig. 8(c), the grain boundary phase is subtly different from Fig. 8(a), which has an incomplete periodic structure with a thickness of about 2 nm. Figure 8(d) shows EDS line profiles for Fe, Sm and V concentrations around the grain boundary in Fig. 8(c). In the grain boundary phase, Sm is enriched up to 36.7 at%, whereas Fe and V are reduced to 52.2 at% and 11.1 at%, respectively. The composition ratio of Sm and (Fe + V) is approximately 1:1.7.
3.4 DiscussionThis study is focusing on the relationship between the coercivity and the microstructure, particularly the grain boundary phase between the ThMn12-type grains of Sm–Fe–M (M = Ti and V) sintered magnets. Note that the coercivity and the microstructure depend on the Sm content of the sintered magnet. The Sm8.6Fe76.5V14.9 magnet has a little coercivity of 44.8 kAm−1, in which the α-(Fe, V) phase formed. This may be a lack of Sm whose content is less than that of the stoichiometry of the ThMn12-type compound effectively because O and C contamination during the powder-metallurgy process with lubricant consumed the Sm element to form the Sm oxycarbide phase. Therefore, it suggests that the Sm8.6Fe76.5V14.9 magnet has no “Sm-rich” grain boundary phases so that it has a little coercivity. The Sm10.5Fe74.1V15.4 magnet exhibited the high coercivity of 648.7 kAm−1 formed Sm-rich phase not only at the triple junctions but also at the grain boundary between the ThMn12-type grains. Meanwhile, the Sm10.0Fe81.7Ti8.3 magnet has no coercivity, although the Sm content is almost the same as that of the Sm10.5Fe74.1V15.4 magnet which exhibits the high coercivity. The microstructural observation revealed that the Sm(Fe, Ti)2 phase to be the liquid phase at sintering temperature was segregated at the triple junctions, not at the grain boundary. The effect of the grain boundary phase on the coercivity has also been shown as the report by Ogawa et al.3) Judging from these results, the grain boundary phase surrounding the ThMn12-type grains could affect the occurrence of the coercivity.
It is noticed that the Sm10.5Fe74.1V15.4 magnet formed grain boundary phases between the ThMn12-type grains, but the Sm10.0Fe81.7Ti8.3 magnet did not. The Sm(Fe, Ti)2 phase was observed in the Sm10.0Fe81.7Ti8.3 magnet at only the triple junctions as shown in Fig. 5(c). On the other hand, the Sm10.5Fe74.1V15.4 magnet formed not only Sm(Fe, V)2 phase but also the Sm-rich phase at the triple junctions as shown in Fig. 6(c). The Sm(Fe, V)2 phase and the Sm-rich phase could cause the eutectic reaction and to lower the melting point of them. The low melting point of the eutectic reaction would be the reason why the Sm10.5Fe74.1V15.4 magnet formed the grain boundary phase as shown in Fig. 8. In order to form a grain boundary phase with suitable morphology, it is important to contain the Sm-rich phase in addition to the Sm(Fe, M)2 phase in order to cause the eutectic reaction at relatively low temperatures.
Such a phase equilibria can be obtained by selecting appropriate stabilizing elements M and adjusting to Sm-rich off-stoichiometric compositions.
Magnetic properties and microstructures of Sm–Fe–M (M = Ti, V) anisotropic sintered magnets with the ThMn12-type structure prepared by the powder-metallurgy were investigated in this study. Sm–Fe–Ti sintered magnets show a low level of coercivities of about 10 kAm−1 regardless of Sm content up to 10.0 at%. Through microstructural observation of the Sm10.0Fe81.7Ti8.3 magnet, Sm(Fe, Ti)2 and Sm3(Fe, Ti)29 phases were observed at the triple junctions, but no grain boundary phases were observed at the grain boundary between the ThMn12-type grains. On the other hand, Sm–Fe–V sintered magnets with low Sm content, such as Sm8.6Fe76.5V14.9 and Sm9.6Fe75.6V14.8, have little coercivities of 44.8 kAm−1 and 58.3 kAm−1, whereas the magnet with higher Sm content, such as Sm9.9Fe75.2V14.9 and Sm10.5Fe74.1V15.4 exhibit high coercivities of 670.3 kAm−1 and 648.7 kAm−1 with remanences of 0.60 T and 0.73 T, respectively. (BH)max reaches 97.6 kJm−3 for the Sm10.5Fe74.1V15.4 magnet, and hysteresis loops measured along the parallel and perpendicular directions to the alignment field show that it is clearly an anisotropic sintered magnet. At the grain boundary of the Sm10.5Fe74.1V15.4 magnet, two types of the phases with different crystal structures, compositions and thickness are observed by STEM-EDS analysis; one is an amorphous structure with a composition ratio of Sm:(Fe + V) = 1:1 with a thickness of about 1 nm, another is an incomplete periodic structure with a composition ratio of Sm:(Fe + V) = 1:1.7 with the thickness of about 2 nm. This study suggests that the inducement of the grain boundary phase formed with Sm-rich off-stoichiometric compositions, and lowering the melting point of the grain boundary phase by selecting suitable stabilizing elements M, may be important to enhance the coercivity of the ThMn12-type magnets.
