Abstract
Nd12.3−xDyxFe81.7Zr0.8Nb0.8Cu0.4B6.0 ribbons have been prepared by melt-spun at 30 m/s and subsequent annealing. The influence of addition of Dy on the crystallization behavior, magnetic properties, and microstructure were investigated. Differential scanning calorimeter (DSC) and X-ray diffraction (XRD) revealed a single-phase material. Microstructure studies using transmission electron microscopy (TEM) had shown a significant microstructure refinement with Dy addition. Wohlfarth’s analysis showed that the exchange coupling interactions increased first with Dy content x increasing, reached the maximum value at , and then slightly decreased with x further increasing. Optimal magnetic properties with T, kA/m, and kJ/m3 are achieved by annealing the melt-spun ribbons with at% at 700°C for 10 min.
1. Introduction
Nd-Fe-B magnets markets have been growing [1] and nanocrystalline Nd-Fe-B has been developed as an essential permanent magnet material for industrial applications [2]. The nanostructure and the composition of the alloy have an influence on the magnetic properties of the isotropic nanocrystalline Nd-Fe-B permanent magnets; extensive efforts have been made via some special methods [3, 4] or adjusting composition to improve the magnetic properties. At present, elemental substitution has been found to be one of the most effective methods for excellent performance. Pr, Zr, Nb, Cu, and Dy substitutions have been widely used in Nd-Fe-B magnets. Studies indicate that Zr or Nd addition can inhibit grain growth, leading to much finer microstructure and higher magnetic properties [5–7]. Cu and Nb addition also can extend the temperature range of the heat-treatment for optimizing the magnetic properties [8]. As we know that the magnetocrystalline anisotropy field () of Dy2Fe14B ( kA/m) is much higher than the one of Nd2Fe14B ( kA/m) [9], Dy substitutes Nd can improve the magnetocrystalline anisotropy field () and optimize the microstructure of the Nd2Fe14B matrix and decrease the grain size [10]. The researchers in the past have been concentrating on nanocomposite Fe3B/Nd2Fe14B or Nd2Fe14B/-Fe [11, 12]. However, the report of the single addition of Dy was rare. In this work, we present our research about the crystallization behavior, microstructure, magnetic properties, and exchange coupling of single phase Nd12.3−xDyxFe79.7Zr0.8Nb0.8Cu0.4B6.0 ( = 0–2.5) ribbons.
2. Experiment
The Nd12.3−xDyxFe79.7Zr0.8Nb0.8Cu0.4B6.0 ( = 0–2.5) ribbons were obtained by applying the melt-spinning technique at a wheel speed of 30 m/s. The ribbon samples with a width of 2-3 mm and a thickness of 40–50 μm were isothermally annealed at 600–800°C for 10 min in vacuum-sealed quartz tubes to crystallize. Phase analysis of the samples was characterized by X-ray diffractometer with Cu-K radiation. The crystallization behavior was examined by differential scanning calorimetry (DSC). Magnetic measurement at room temperature was employed using a LDJ9600 vibrating sample magnetometer (VSM). The length direction of the ribbons was parallel to the applied field in order to minimize demagnetization effect. The microstructures were observed by high-resolution transmission electron microscopy (HR-TEM JEM2010) with thin foils prepared by ion-beam thinning.
3. Results and Discussion
3.1. Phase Analysis
Figure 1 shows the XRD patterns of at different condition. Figure 1(a) shows the alloy has partial crystallization and highlights the weak characteristic peaks of Nd2Fe14B. This indicates that those ribbons still have many amorphous phases and the grain size of the crystalline phase is small. Figure 1(b) shows XRD pattern of as-cast Nd10.8Dy1.5Fe79.7Zr0.8Nb0.8Cu0.4B6.0 alloy annealed at 1050°C for 10 h. The only crystalline phase in the matrix is 2 : 14 : 1 phase at both the annealed condition and as-spun condition. After the annealing treatment, the Nd2Fe14B grains were completely formed, and there is no new phase appearing in the grain boundary. This result provides the possibility of the magnetic exchange coupling interaction between the adjacent grains.
(a)
(b)
Figure 2 shows the differential scanning calorimetry (DSC) curves of the as-spun ribbons with different Dy content at a heating rate of 10°C/min from room temperature to 800°C and the thermal properties determined from these DSC thermal scans are summarized in Table 1. Samples with at% and without other elements of Nd12.3Fe81.7B6.0 show one exothermic peak corresponding to the transformation from amorphous phase to 2 : 14 : 1 phase. It is in accordance with the XRD result in Figure 1. Compared with the curve of Nd12.3Fe81.7B6.0, the ribbon with substitution of 1.5 at% Table 1 and Figure 1 show that Dy and other additional elements retarded the onset temperature of crystallization and peak temperature of crystallization. When the temperature reaches 580°C, the alloy with additional elements has no crystallization. The contrary crystallization phenomenon of improving crystallization tendency of the Mn substitution in Nd2Fe14B/-Fe nanocomposites was also observed by Xie et al. [13].
(a)
(b)
3.2. Effect of Dy Content on Magnetic Properties of Ribbons
Since the amorphous phase is undesired, in order to achieve the best magnetic properties for the samples with different Dy content, a thermal treatment at 600–800°C for 10 min is employed individually to quenched ribbons. Figure 3 summarizes variation of , , and with Dy content. The intrinsic coercivity significantly increases with increasing Dy content. That is because the magnetocrystalline anisotropy field () of the Dy2Fe14B (12600 kA/m) is larger than that of the Nd2Fe14B (5600 kA/m) and they are totally mutual soluble [9]. The remanence of Dy from 0 at% to 0.5 at% doped samples shows little change. Then, the decreases with increasing Dy content. On the one hand, the saturation magnetization () of Dy2Fe14B (0.71 T) is smaller than that of the Nd2Fe14B (1.60 T) [14]. As we all know that Nd and Dy are both rare elements and chemical properties, the atomic radius of Dy is larger than that of Nd. When Dy atoms substitute the site of Nd atoms, the volume and the local atomic structure of Nd will be influenced. With further Dy addition, the effect of Dy on reducing and becomes stronger and the remanence decreases. Magnetic energy product increases initially and then decreases and reaches the peak value when the content of Dy is 0.5 at%. When the content of Dy is more than 1.0 at%, begins to greatly reduce, due to the deviation of demagnetization curve square degree.
(a)
(b)
(c)
The optimal magnetic properties were indicated in Table 2. As can be seen, the addition of Dy can improve significantly the magnetic properties. A small amount of Dy can refine the grains and optimize the microstructure. After the optimum annealing treatment, the coercivity of Nd10.8Dy1.5Fe79.7Nb0.8Cu0.4B6.0 is enhanced from 775.5 kA/m for to 1440 kA/m for . However, with the increase of the Dy content, the remanence is increased first from 0.98 T for to 1.09 T and then decreased to 0.98 T for . The maximum energy product is increased firstly and then decreased. All the maximum energy products of the annealed sample with Dy substitution are larger than 130 kJ/m3. Optimum magnetic properties with T, kA/m and = 169.5 kJ/m3 are achieved by annealing the melt-spun ribbons with at% at 700°C for 10 min. Compared to the Dy-free sample, , , and of alloy with at% are increased by 40%, 11%, and 27%.
3.3. Effect of Dy Content on Microstructure
Figure 4 shows TEM images of the Dy-free and Dy-containing ribbons annealed at 700°C for 10 min. The average grain size is estimated to be 40 to 60 nm for the Dy-free ribbons and the grain size of the Dy-doped ribbons is significantly finer. The average grain size for the samples with is estimated to be 30 to 50 nm, and the grains become more uniformly distributed. It implies that the slight amount of Dy is effective to uniform the grains and suppress the growth of the grains during the heat treatment. The fine and homogeneous microstructure is favored for exchange coupling interaction.
(a)
(b)
Figure 5 shows the HR-TEM images of Nd12.3−xDyxFe79.7Zr0.8Nb0.8Cu0.4B6.0 annealed at 700°C for 10 min. The special morphology phase (large grains are mixed up with small grains) is 2 : 14 : 1 phase and the crystallization is perfectly complete. No boundary phase is found between the grains. The grain boundaries are totally crystallographically coherent. This is different from the structure of the sintered NdFeB magnets, whose coercivity dramatically depends on the Nd-rich phase between and around the grains. In order to further analyze the large grains sandwiched small grains inside the grains and on the grain boundaries, the Fourier filtered images corresponding to different sections of Figure 5 are shown in Figure 6. As can be seen, either inside the grains or on the grain boundaries, the lattice regular arrangement and the grain boundaries are partly coherent. No second phase or grain boundary phase is found. It also proves that the coercivity mechanism in this series of materials is not similar to the traditional one in the sintered magnets. The magnetic coercivity is probably realized by the effect of the magnetic exchange-coupling interaction between grains.
(a)
(b)
(a)
(b)
(c)
The exchange coupling interaction is often evaluated using plots [15], which can be defined as , in which is reduced magnetization and is reduced remanence with respect to the applied magnetic field, . According to Wohlfarth’s analysis [16], higher positive peaks indicate stronger exchange coupling interactions among grains. Figure 7 shows plots of for the annealing Nd12.3−xDyxFe79.7Zr0.8Nb0.8Cu0.4B6.0 ( = 0–2.0) samples. All samples are composed of one magnetically hard phase, and the remanence ratio is larger than 0.5, which indicates the existence of the exchange coupling between grains. It is clear that the magnitude of peaks increases first with Dy content increasing and reaches maximum value at and then decreases with further increase of Dy content, indicating that the intergranular exchange coupling effect appears to strengthen with the increase of up to 0.5 and then diminish with further increase in . The plots of for Nd10.3Dy2Fe79.7Zr0.8Nb0.8Cu0.4B6.0 () sample has still a high peak. According to Wohlfarth’s analysis, higher positive peaks indicate stronger exchange coupling interactions among grains. The XRD pattern in Figure 1(b) shows that there is only one crystalline phase (Nd2Fe14B). It can imply that has a strong exchange coupling effect between Nd2Fe14B grains. The cure of at% has a wide curve peak in the positive axis. The trend is a change in the magnetic energy with Dy content. The exchange coupling interactions increased fist with Dy content increasing, reached the maximum value at , and then slightly decreased with further increasing. The magnetic properties and it is consistent. Hence, the changes in magnetic properties of the samples on increasing Dy content may mainly arise from the variation of the exchange coupling interaction in the samples.
4. Conclusion
Magnetic properties of nanocrystalline Nd12.3Fe81.7Zr0.8Nb0.8Cu0.4B6.0 ribbons prepared by melt spinning and subsequent annealing were significantly enhanced by the substitution of Dy for Nd. The main results obtained are the following:(1)The slight content of Dy stabilized the amorphous phase and enhanced the thermal stability of amorphous phase in the melt-spun.(2)The intrinsic coercivity and maximum energy product increased with increased with increasing Dy content and then decreased with further increasing Dy content. However, the remanence polarization decreased with Dy content monotonically. The excellent magnetic properties with = 169.5 kJ/m3, = 809.2 kA/m, and = 1.02 T were obtained by annealing a melt-spun amorphous Nd11.8Dy0.5Fe81.7Zr0.8Nb0.8Cu0.4B6.0 alloy at 700°C for 10 min.(3)The exchange coupling interactions increased first with Dy content increasing, reached the maximum value at , and then slightly decreased with further increasing. In all samples, they have a strong exchange coupling between neighboring grains. The fine and homogeneous microstructure is favored for exchange coupling interaction.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgment
This study is supported by the National Natural Science Foundation of China under Grant no. 51174030.