Tunable Curie temperature around room temperature and magnetocaloric effect in ternary Ce–Fe–B amorphous ribbons

The change of magnetic entropy in magnetic materials is accompanied by an exothermic or endothermic process, which is regarded as a magnetocaloric effect and can be used for refrigeration [1–3]. Compared with conventional refrigeration, magnetic refrigeration is more energy-saving and environmentally friendly. The abrupt change of magnetization accompanied by the structural or volume transition leads to a large entropy change and large refrigeration capacity (RC) in first-order phase transition for crystalline magnets [4–7]. In amorphous magnets, the atomic arrangement is in long-range disorder, and so there is neither volume nor structural change accompanying the magnetic phase transition [8, 9]. Though their magnetization change is slow with a low entropy change in the second-order phase transition, amorphous magnets are also attractive for magnetic refrigerant applications due to their wide working temperature range, reduced magnetic hysteresis, and good mechanical properties [8, 9].

The Fe atom bears the largest magnetic moments per unit mass, which is desirable for obtaining large entropy change and refrigeration capacity, but Curie temperatures are much higher than room temperature in Fe-based amorphous alloys. The doping of refractory metals Nb, Zr, Ti, and Cr can not only tune Curie temperature but also improve the glass-forming ability [10–14]. Meanwhile, the saturation magnetization and RC are reduced due to the nonmagnetic element additions. Fe and rare-earth elements can form amorphous alloys, which can be used as magnetic refrigerants [15–18]. In this work, low-cost rare-earth Ce–Fe–B amorphous alloys are prepared by Ce, Fe, and FeB using a melt-spinning method. In these Ce–Fe–B magnets, the magnetocrystalline anisotropy is negligible at low temperature and the Curie temperature is around room temperature. The magnetic phase transition and magnetocaloric effect are systematically investigated in these ternary Ce–Fe–B amorphous magnets with the variation of Ce/Fe atomic ratio.

The precursor ingots with nominal composition of Ce_13−xFe_81+xB₆ (x  =  0, 0.5, 1, 1.5, and 2) were produced from the mixture of high-purity Ce, Fe, and Fe–B alloys by arc-melting method under an argon gas. Melting was performed at least three times to ensure a good homogeneity of composition. An excess of 2wt.% for Ce was used to compensate the loss of volatilization during the melting process. The ribbons were obtained directly by induction-melting the precursor ingot and then ejecting the melt onto the surface of a rotating copper wheel with a surface velocity of 50 m s⁻¹. The x-ray diffraction (XRD) was used to check the phase constitution. Magnetic measurements were performed using a superconducting quantum interference device vibrating sample magnetometer with the field direction in the plane of the ribbons, and therefore the demagnetization factor was neglected.

Figure 1 shows the XRD patterns of these ribbons. There exist much broadened diffraction peaks with weak intensity, which should be ascribed to the amorphous structure with atomic arrangement in short range order [16]. Since the angles of diffraction peaks correspond well to those of Re₂Fe₁₄B, there may be a small amount of nanocrystalline phase embedded in the amorphous matrix. For a comparison, the XRD pattern is also shown in figure 1 for Ce₁₂Fe₈₂B₆ ribbons prepared with low wheel-surface velocity of 20 m s⁻¹. The characteristic diffraction peaks are very strong, indicating long range order for atomic arrangement in crystalline structure. The atomic arrangement in amorphous structure may significantly affect the magnetic structure, which could be checked by the magnetization dependence on temperature. We first lowered the temperature to 10 K in the absence of an applied magnetic field. Then we increased the temperature from 10 K to 400 K under a field of 0.2 kOe and recorded the value of magnetic moments, which is called magnetization variation with temperature in zero-field cooling (ZFC). After that, we recorded the magnetization as the temperature decreased from 400 K to 10 K under a field of 0.2 kOe in field cooling (FC). As shown in figure 2, with the decrease of temperature the magnetization increases monotonously near room temperature, suggesting the occurrence of a magnetic phase transition from paramagnetic to ferromagnetic state [18, 19]. The slow increase in magnetization is attributed to the fluctuation of the exchange integral caused by the structural disorder in amorphous alloys [15, 20]. The curves of FC and ZFC nearly overlap in the whole temperature range, indicating negligible magnetic hysteresis and thermal hysteresis for these amorphous magnets. Curie temperature, determined by the maximum of $\text{d}M/\text{d}T$ [8], is in the range between 270 K and 300 K (shown in figure 3), and is tunable by the variation of the Ce/Fe atomic ratio. The Curie temperature is about 460 K in Ce–Fe–B ribbons prepared with the copper surface velocity of 27 m s⁻¹, which is owing to the long-range order for atomic arrangement in the Ce₂Fe₁₄B crystalline structure [21]. However, for the amorphous ribbons prepared with higher quenching rate, the phase- transition temperature decreases to room temperature.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Curie temperature is strongly dependent on the Fe–Fe atomic distance and coordination number [8, 10, 13, 22]. It has been argued that the number of surrounding Fe atoms for each Fe site plays a crucial role in determining ${{T}_{\text{C}}}$ for Fe–B amorphous alloys [8, 13]. In these Ce–Fe–B amorphous magnets, the increase of Fe and the corresponding decrease of Ce atomic percent would lead to the increase of the coordination number ${{Z}_{\text{Fe}-\text{Fe}}}$. Hence, like the Fe–Zr and Fe–Ti alloys, it can be expected that the increase of the number of nearest neighbor Fe atoms is the origin of the systematic increase of ${{T}_{\text{C}}}$ in these amorphous magnets [13, 22].

The phase transition is accompanied by an endothermic process, which depends on the entropy change from the magnetic ordering to disordering state. To determine the value of entropy change, the magnetization isotherms were measured under applied field up to 5 T. Figures 4(a)–(c) show the magnetization isotherms of Ce_13−xFe_81+xB₆ amorphous ribbons for x  =  0, 1, and 2, respectively. In rare-earth magnets the magnetic structure is complicated and includes helical structures and the spin configuration of the antiferromagnet. It can be seen that the magnetization reaches saturation quickly at the low temperature of 150 K, which is similar to that in the ferromagnet based on Gd in Gd₆₀M₃₀In₁₀ (M  =  Co, Ni, and Cu) at 5 K [15]. In spin glass alloys, it is hard to saturate even at the applied field of 5 T [10, 15, 19]. This fact verifies the ferromagnetic and excellent magnetically soft properties in these Ce–Fe–B amorphous magnets. Ce–Fe–B bears magnetocrystalline anisotropy of 26 kOe in a Ce₂Fe₁₄B crystalline structure [23], but the anisotropy is negligible in an amorphous structure. It is also noted that the magnetization at 150 K for Ce₁₁Fe₈₃B₆ is greater than that of Ce₁₃Fe₈₁B₆ and Ce₁₂Fe₈₂B₆ ribbons, which should be attributed to the higher content of Fe in Ce₁₁Fe₈₃B₆ ribbons.

Zoom In Zoom Out Reset image size

To check the character of phase transition in the Ce–Fe–B amorphous magnets, the Arrott plots are shown in figures 5(a)–(c), which are derived from $M$–$H$ isotherms. According to the Banerjee criterion [24], the positive slope on the Arrott curve confirms the typical characteristic of a second-order magnetic transition from the ferromagnetic to paramagnetic state. Based on the thermodynamic theory, the magnetic entropy change is calculated from the isothermal magnetization by the Maxwell relation, $\Delta {{S}_{\text{M}}}(T,H)={\int}_{0}^{H}{{\left(\partial M/\partial T\right)}_{H}}\text{d}H$ [25]. Figure 6 shows the temperature dependence of entropy change ($- \Delta {{S}_{\text{M}}}$) for different applied field changes. Compared with those of many crystal magnets, though the maximum of ($- \Delta {{S}_{\text{M}}}$) is smaller, a very broad distribution of ($- \Delta {{S}_{\text{M}}}$) is exhibited, indicating a wide temperature span which is desirable for an Ericson cycle magnetic refrigerator. The wide temperature span results from the slow magnetization change in the magnetic phase transition due to the disordered structure [8, 9]. The maximum of the entropy change ($- \Delta {{S}_{\text{M}}}$) for a field change of 0–5 T increases from 3.3 to 3.9 J kg⁻¹ K, with x increasing from 0 to 2 in Ce_13−xFe_81+xB₆. The increase of Fe content would lead to an increase of saturation magnetization and a large change of magnetization in magnetic phase transition, which should be ascribed to the variation of entropy change ($- \Delta {{S}_{\text{M}}}$) with the Ce/Fe atomic ratio. With the increase of Fe content, the corresponding position of the maximum of ($- \Delta {{S}_{\text{M}}}$) shifts to higher temperature. This fact is consistent with the variation of Curie temperature with Ce/Fe atomic ratio, indicating the tunable working temperature for magnetic refrigeration in these Ce–Fe–B amorphous magnets.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

For the second-order magnetic phase transition, the field dependence of the maximum magnetic entropy change can be expressed as $\left(- \Delta {{S}_{\text{M}}}\right)\propto {{H}^{n}}$ based on the mean field theory [26, 27]. The relations of ($- \Delta {{S}_{\text{M}}}$) maximum and $H$ are shown in figure 7. As shown in the fitting figure plotted in the inset, the value of $n$ is 0.73, which is consistent with the results for Fe-based amorphous alloys [27] but higher than the theoretical value 2/3 for mean field theory. Possibly it is the failure of the mean field picture for these second-order phase-transition materials [28]. In addition, the value of $n$ is also affected by the structural inhomogeneity due to the free volume, atomic cluster, and even nano-crystallized grains embedded in the disordered matrix [29].

Zoom In Zoom Out Reset image size

In the refrigeration cycle, accompanied by the entropy change, the working material absorbs the heat from the load and the RC can be quantified by the heat absorbed per kilogram [9]. As suggested by Wood and Potter [30], RC is determined by the integral $\text{RC}={\int}_{{{T}_{\text{cold}}}}^{{{T}_{\text{hot}}}} \Delta {{S}_{\left(T, \Delta H\right)}}\text{d}T$. Here, ${{T}_{\text{hot}}}$ and ${{T}_{\text{cold}}}$ are defined as the temperature boundary at the hot and cold ends, respectively, corresponding to the half maximum of ($- \Delta {{S}_{\text{M}}}$) on the $\left(- \Delta {{S}_{\text{M}}}\right)-T$ curve. The values of RC for Ce–Fe–B amorphous magnets are 117–150 J kg⁻¹ and 319–420 J kg⁻¹, respectively, for magnetic field changes of 0–2 T and 0–5 T, which are larger than those of most amorphous alloys with ${{T}_{\text{C}}}$ near room temperature. Though possessing small magnetic entropy change, their RCs (for magnetic field change 0–5 T) are also higher than those of most crystalline materials for room-temperature refrigeration, such as La_0.8Ce_0.2Fe_11.5Si_1.5C_0.4 (330 J kg⁻¹) [31], Gd₅Ge₂Si₂ (200 J kg⁻¹) [32], and MnFeP_0.45As_0.35 (225 J kg⁻¹) [6, 33]—though, of course, considering the too-large temperature span, the real refrigerant capacity may be reduced more or less in practical technology for these amorphous magnets. Fe bears the largest magnetic moments per kilogram among the magnetic elements. The high RC should be attributed to the higher content of the magnetic element Fe with fewer amounts of nonmagnetic elements as well as the amorphous structure. Table 1 lists the magnetic properties (the values of ${{T}_{\text{C}}}$, ($- \Delta {{S}_{\text{M}}}$), and RC) for some typical magnets as well as Ce–Fe–B amorphous magnets for room-temperature refrigeration. Ce is cheap and the most abundant of rare earth resources and overstocked due to the overconsumption of Nd and Pr elements [23]. Combining the particular merits coming from their amorphous nature and large RC, Ce–Fe–B amorphous magnets are also attractive candidates for use as magnetic refrigerants in room- temperature refrigeration.

In summary, ternary Ce–Fe–B amorphous magnets can be prepared by the melt-spin method with higher quenching rate. Compared to Ce–Fe–B crystalline magnets, the Curie temperature in these amorphous magnets decreases to room temperature and magnetocrystalline anisotropy is negligible at low temperature. The working temperature range is wide, and the magnetic as well as thermal hysteresis is negligible due to the disordered atomic arrangement. The entropy change that accompanies the magnetic phase transition is larger than those of most amorphous magnets for room-temperature refrigeration, which should be attributed to the higher content of the magnetic element Fe and lower content of the nonmagnetic elements. The values of RC for Ce–Fe–B amorphous magnets are 117–150 J kg⁻¹ and 319–420 J kg⁻¹, respectively, for magnetic field changes of 0–2 T and 0–5 T, which are larger than those of most amorphous alloys with ${{T}_{\text{C}}}$ near room temperature. Due to the considerable refrigeration capacity, low cost Ce–Fe–B magnets are also attractive candidates for use as magnetic refrigerants in room-temperature refrigeration.

The present work was supported by the National Basic Research Program of China (Grant No. 2014CB643702) and the National Natural Science Foundation of China (Grant Nos. 51461033, 11274357, 51590880).