Project supported by the National Natural Science Foundation of China (Grant No. 51901052) and the Science Foundation of Guangxi Zhuang Autonomous Region, China (Grant No. 2018GXNSFAA281294).
Project supported by the National Natural Science Foundation of China (Grant No. 51901052) and the Science Foundation of Guangxi Zhuang Autonomous Region, China (Grant No. 2018GXNSFAA281294).
† Corresponding author. E-mail:
Project supported by the National Natural Science Foundation of China (Grant No. 51901052) and the Science Foundation of Guangxi Zhuang Autonomous Region, China (Grant No. 2018GXNSFAA281294).
The crystal structures, magnetization, and spontaneous magnetostriction of ferromagnetic Laves phase Pr1 − xTbxFe1.9 compounds are investigated in a temperature range between 5 K and 300 K. High resolution synchrotron x-ray diffraction (XRD) analysis shows that different proportions of Tb in Pr1 − xTbxFe1.9 alloys can result in different easy magnetization directions (EMD) below 70 K, i.e., [100] with x = 0.0, and [111] with x ≥ 0.1. This indicates Tb substitution can lead the EMD to change from [100] to [111] with x rising from 0.0 up to 0.1. The Tb substitution for Pr reduces the saturation magnetization Ms and the magnetostriction to their minimum value when x = 0.6, but it can increase low-field (0 ≤ H ≤ 9 kOe, the unit 1 Oe = 79.5775 A · m−1) magnetostriction when x = 0.8 and 1.0 at 5 K. This can be attributed to the larger magnetostriction of PrFe1.9 than that of TbFe1.9, as well as the decrease of the resulting anisotropy due to Tb substitution at low temperatures.
The binary C15 cubic Laves phase RFe2 (R = rare-earth elements) compounds are known to exhibit giant magnetostriction, which has been widely used as magnetostrictive materials in transducers, spintronics, information storage, sensors, and the design of novel electronic devices.[1–9] Especially, the low-cost light rare earth PrFe2 alloy, which has a large calculated magnetostriction (λ111 ∼ 5600 ppm at 0 K) according to single ion model theory, has received much attention in recent years.[10–14] However, this alloy shows high magnetocrystalline anisotropy due to the strong anisotropic 4f shells, and thus it possesses high switching fields at low temperatures.[15] In the past few years, much attention has been paid to developing the anisotropy compensating system similar to the well-known Terfenol-D, by combining two different RFe2 terminals with opposite signs of anisotropy constant, e.g., Tb0.2Dy0.8−xPrx(Fe0.9B0.1)1.93, TbxDy1 − xPr0.3(Fe0.9B0.1)1.93, and Tb0.1Ho0.9−xPrx(Fe0.9B0.1)2 systems, etc.[11,16,17] Among them, Pr1−xTbxFe1.9 is a promising anisotropy compensating system, while many Tb alloys have favorable characteristics at low magnetic fields.[18–23] Since the easy magnetic direction (EMD) of PrFe2 lies along [100], while the easy magnetic direction of TbFe2 lies along [111] at low temperatures,[1,15] their magnetocrystalline anisotropy constants with different signs may be offset by each other.[24] In addition, TbFe2 has large magnetostriction (λ111 ∼ 4400 ppm at 0 K) and the same sign of the magnetostriction as PrFe2, which might enlarge the low-field magnetostriction in the Pr1−xTbxFe1.9 system with appropriate Tb component. Furthermore, according to the x-ray diffraction studies on PrFe2, its easy magnetic direction lies along [111] above its spin reorientation temperature Tsr (about 70 K), while it lies along [100] when T < Tsr.[15] This transition is considered as a morphotropic phase boundary (MPB) phenomenon in ferromagnets, and it has driven the design of materials with giant low-field magnetostriction.[25–28] Nevertheless, most of this meaningful research mainly focuses on the heavy rare-earth system. Further MPB study on the low-cost light rare-earth containing systems is still needed, for example, the Pr1−xTbxFe1.9 system, in which PrFe1.9 and TbFe1.9 are the two starting components with different crystal structures.
The objective of the presented work is to study the effect of the rare-earth (Tb) substitution for Pr on the structural symmetry, spin reorientation temperature, magnetization, and magnetostriction in the ferromagnetic Pr1−xTbxFe1.9 alloys.
Ingots with Pr1 − xTbxFe1.9 (x = 0.0, 0.05, 0.1, 0.2, 0.4, 0.6, 0.8, 1.0) stoichiometry were prepared by melting the high purity metals in a magneto-controlled arc furnace in an argon atmosphere.[12,23] The alloys were prepared from materials with the following purities: Pr and Tb (99.9 wt%), Fe (99.8 wt%). The ingots each with about 1 g were pressed into disks and wrapped with tantalum foils, and then loaded into a cylindrical graphite pipe heater. The assembly was pressed to 6 GPa by a hexahedral anvil press and heated to 1173 K for 30 min. Conventional x-ray diffraction (XRD) analysis with using Cu-Kα radiation with a Rigaku D/Max-ga diffractometer was made. The samples for XRD were ground into powders to reduce the preferred orientation effect. The powder XRD patterns for Pr1−xTbxFe1.9 alloys with different Tb concentrations at room temperature were obtained and are shown in Fig.
It is found that all the polycrystalline alloys consist predominantly of cubic Laves phase with MgCu2-type structure. It is generally believed that the cubic Laves phase could not be obtained under normal pressure in high-Pr alloys due to the large radius of Pr3+. Therefore, the formation of the Laves phase in high-Pr alloy should be ascribed to the effect of the high-pressure annealing method.[12,23] There is a hint of impurity phase, i.e., hcp-(Pr, Tb) coexisting in the alloys,[22,23] which is marked in the figure. High precision step scanning for 40.5° ≤ 2θ ≤ 43° and 70.5° ≤ 2θ ≤ 72.5° enabled the spontaneous magnetostriction-induced splitting of the {222} and {440} lines to be investigated, respectively. The XRD peaks were fitted by Jade 6.5 XRD analytical software, and the profiles of the {440} and {222} lines of Kα2 were deducted with a standard method. Next, superconducting quantum interference device magnetometer (SQUID) was used to measure the magnetization curves of the samples. The temperature dependence of magnetization was also measured by using SQUID to determine the spin reorientation temperature Tsr.
The magnetostriction was then measured by a standard strain-gauge technique in the parallel (λ||) direction and perpendicular (λ⊥) direction of magnetic field for each sample (a cuboid block with a size of about 6 mm× 4 mm× 1 mm for each sample), with the magnetic field supplied by “Quantum Design” (PPMS).
The curves of magnetostriction λ|| versus applied magnetic field H for various x values at 5 K are measured and shown in Fig.
Figure
The ferromagnetic transition can involve structural change.[19] In order to verify the possible change of EMD in such transitions, as well as the influence of Tb on the crystal structure in the alloys, a high-precision XRD step scanning is performed in the alloy with x = 0.1 under cooling condition, which is shown in Fig.
To further investigate the effect of Tb substitution for Pr on the magnetizations and the magnetocrystalline anisotropy, the magnetization curves of Pr1−xTbxFe1.9 are measured at 5 K and 300 K, which are shown in Figs.
As for the magnetization result at 300 K, the magnetization decreases with Tb concentration increasing till x = 0.6 and increases when x ≥ 0.6, which is similar to the result at 5 K. The alloys with x = 0.0 possess the maximum M/M(70 kOe) value as shown in the inset of Fig.
In conclusion, the substitution of Tb for Pr leads magnetostriction and saturation magnetization Ms to reach their corresponding minimum values when x = 0.6, but it also increases the value of the magnetostriction in a magnetic field range of 0 ≤ H ≤ 9 kOe and reduces the magnetocrystalline anisotropy when x = 0.8 and x = 1.0 at 5 K. The easy magnetization direction (EMD) lies along the [100] direction in a temperature range between 15 K to 70 K when x = 0.0, but lies along the [111] direction when x ≥ 0.1. This indicates that the Tb substitution with x ≥ 0.1 can lead the EMD to change from [100] to [111].
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] |