Energy efficiency and environmentally friendliness have become the most urgent requirement of refrigeration technology due to the increasing energy shortage and environmental deterioration. Therefore, efforts have been made to develop new climate-friendly refrigeration solutions with high efficiency. The magnetic refrigeration (MR) technology, by using the endothermic and exothermic characteristic of a refrigerating medium upon the variation of the magnetic field, has displayed promising potential due to its compactness, high efficiency, energy saving, and environmental harmlessness.^[1–5]

As the working materials play a critical role in the efficiency of the magnetic refrigerator, numerous magnetic alloys that exhibit fine magneto-caloric effect (MCE) have been studied intensively in recent years.^[5–25] Amongst these MCE materials, the amorphous alloys display better mechanical properties and corrosion resistance, in addition to their excellent magnetic proprieties. For instance, the Gd-based amorphous alloys show considerable prospects as magnetic refrigerants based on their soft magnetic performance with negligible magnetic hysteresis at different temperatures, the ultrahigh refrigeration capacity (RC) corresponding to their broadened magnetic entropy change (−ΔS_m) peak, and especially their tunable Curie temperature (T_C) over a wide range of composition.

However, as the situation in the Gd–Co binary metallic glasses,^[23–25] the glass formability (GFA) and MCE of the Gd-based amorphous alloys become worse with the increasing T_C.^[8–22] Therefore, although the multicomponent Gd-based bulk metallic glasses show high glass formability and outstanding MCE at low temperature, the Gd-based amorphous alloys can only be fabricated into thin ribbons when improving the T_C to near the room temperature and their maximum magnetic entropy change () is usually less than half of the value of pure Gd. So far, the T_C of the Gd-based fully amorphous alloys is lower than 282 K, which makes it difficult to compose the amorphous composites with a table-like −ΔS_m profile suitable for a magnetic refrigeration cycle of a domestic refrigerator or air conditioner. In this work, through adding a small amount of Ni to substitute Gd of the binary Gd₅₀Co₅₀ amorphous alloy, we obtain Co₅₀Gd_{50 − x}Ni_x alloys in an attempt to achieve improved formability and enhanced T_C, without excessive reduction the MCE of the initial alloy. The thermal and magnetic properties of the Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) amorphous alloys were measured. Based on the experimental results obtained from these Ni addition samples, the change of the formability, T_C, as well as the compared to the Gd₅₀Co₅₀ amorphous alloy were investigated. The adiabatic temperature rise (ΔT_ad) was calculated to assess the application perspective as a magnetic refrigerant around room temperature.

Samples with nominal compositions of Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, 3, 4, and 5) were firstly prepared into ingots by arc-melting of pure Gd, Co, and Ni metals (purity ≥ 99.9 at.%) for several times under a Ti-gettered Ar atmosphere, and then quenched into ribbons with a thickness of 30–40 μm by a single roller melt-spinning method. The structural characteristics of the as-spun ribbons were tested by the Rigaku x-ray diffraction (XRD, model D\max-2550) using Cu K_α radiation. The thermal properties of the amorphous ribbons were obtained from the differential scanning calorimetry (DSC) curves measured by a Perkin-Elmer calorimeter (DIAMOND) at a heating rate of 20 K/min. The temperature dependence of magnetization (M–T) curves and the isothermal magnetization (M–H) curves were measured in a Quantum Design physical properties measurement system (PPMS, model 6000), and the magnetic proprieties of Curie temperature (T_C) and magnetic entropy change (−ΔS_m) were calculated from the M–T and M–H curves accordingly.

Figure 1 shows XRD patterns of as-spun Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, 3, 4, and 5) ribbons. The Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) as-spun ribbons display a representative amorphous characteristic with a broad diffraction hump and no obvious crystalline peaks in their XRD patterns. In contrast, the Co₅₀Gd₄₆Ni₄ as-spun ribbon is partially crystalline and the Co₅₀Gd₄₅Ni₅ as-spun ribbon is almost fully crystalline as their XRD patterns appear obvious crystalline peaks of Gd₂Co phase.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. XRD patterns of the as-spun ribbons of Co50Gd50 − xNix (x = 1, 2, 3, 4, and 5).

DSC measurements under a heating rate of 20 K/min were performed to obtain the thermodynamic parameters of the Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) amorphous ribbons. As shown in Fig. 2, the DSC curves of the Co₅₀Gd_{50 − x}Ni_x glassy ribbons exhibit a typical endothermic reaction before the sharp exothermic peaks of crystallization, which further indicates the glassy structure of the alloys. The glass transition temperature T_g, the liquidus temperature T_l, and the crystallization temperature T_x obtained from the DSC curves are listed in Table 1. It is found that the T_g and T_l of the Co₅₀Gd₅₀ amorphous alloy get increased with Ni substituting the Gd elements, which is most likely ascribed to the introduced 3d–3d direct exchange interactions stronger than the indirect interactions from Gd. While the calculated supercooled liquid region ΔT_x (ΔT_x = T_x – T_g) that is used to indicate the thermal stability of the supercooled liquids becomes narrow with Ni addition. It may be related to the reduction of the 3d–4f indirect exchange interactions caused by the replacement of Gd. As is well known, the f–d hybridization stabilizes the supercooled liquid or the glass against crystallization by the localization of the electrons,^[26,27] and as a result, the reduced 3d–4f interactions in the Co₅₀Gd_{50 − x}Ni_x amorphous alloys will weaken the f–d hybridization, thereby broken the stable state of the undercooled liquids and deteriorate their glass-forming ability (GFA). According to the DSC results and the above analysis, the substitution of 1 at.% Ni hardly affects the thermal stability and GFA of the alloy, while the substitution of 2–3 at.% Ni weakens the corresponding thermal stability and GFA, which is well consistent with the results of XRD.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. DSC traces of the Co50Gd50 − xNix (x = 1, 2, and 3) amorphous ribbons obtained at a heating rate of 20 K/min.

Table 1.

Table 1.

Table 1. Thermal, magnetic, and magnetocaloric properties of the Co50Gd50 − xNix (x = 1, 2, and 3) and Gd50Co50 amorphous alloys. . Composition Tg/K Tx/K ΔTx/K TC Co50Gd50 507 557 50 267 Co50Gd49Ni1 529 579 50 273 Co50Gd48Ni2 523 571 48 280 Co50Gd47Ni3 538 558 20 289 Composition /J · kg−1 · K−1 ΔTad/K 1 T 2 T 3 T 4 T 5 T 2 T 5 T Gd50Co50 1.38 2.36 3.16 3.92 4.6 – – Co50Gd49Ni1 1.53 2.53 3.39 4.17 4.90 2.32 4.63 Co50Gd48Ni2 1.34 2.26 3.06 3.78 4.46 – – Co50Gd47Ni3 1.27 2.12 2.85 3.52 4.15 2.10 4.30

Table 1. Thermal, magnetic, and magnetocaloric properties of the Co50Gd50 − xNix (x = 1, 2, and 3) and Gd50Co50 amorphous alloys. .

Figure 3(a) shows the hysteresis loops of the Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) amorphous ribbons measured at 10 K and 350 K under a field of 5 T. The amorphous ribbon is ferromagnetic at 10 K and becomes paramagnetic at room temperature. The saturation magnetization (M_s) is about 166 A·m²/kg for Co₅₀Gd₄₉Ni₁, 159 A·m²/kg for Co₅₀Gd₄₈Ni₂, and 153 A·m²/kg for Co₅₀Gd₄₇Ni₃, respectively. Therefore, the of the Co₅₀Gd₄₉Ni₁ amorphous ribbon is expected to be slightly higher than that of the Co₅₀Gd₄₈Ni₂ and Co₅₀Gd₄₇Ni₃ amorphous ribbons because is closely related to the M_s of the amorphous alloys.

Fig. 3.

Figure Option ViewDownloadNew Window

Fig. 3. (a) The hysteresis loops of the Co50Gd50 − xNix (x = 1, 2, and 3) amorphous ribbons measured at 10 K and 350 K under a field of 5 T. (b) The M–T curves of the Co50Gd50 − xNix (x = 1, 2, and 3) amorphous ribbons under a field of 0.03 T, the inset is μeff of the amorphous alloys. (c) Compositional dependence of TC for the Co50Gd50 − xNix (x = 1, 2, and 3) and Gd50C50 − yNiy (y = 2 and 5) amorphous alloys.

The M–T curves of the Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) metallic glasses were measured to reveal the magnetic transition temperature of the alloys. The Curie temperature T_C, as shown in Fig. 3(b), is about 273 K for Co₅₀Gd₄₉Ni₁, increased to 280 K for Co₅₀Gd₄₈Ni₂, and improved to 289 K for Co₅₀Gd₄₇Ni₃. Therefore, through adding Ni to the binary Gd₅₀Co₅₀ amorphous alloy to reduce the Gd content, the Curie temperature of the Co₅₀Gd_{50 − x}Ni_x amorphous alloys is successfully improved to a higher temperature, especially, T_C of the Co₅₀Gd₄₇Ni₃ amorphous ribbon is higher than that of most Gd-based amorphous alloys.^[8–25]

In order to investigate the mechanism for the improvement of the magnetic transition temperature of the Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) amorphous alloys, we constructed the compositional dependence of T_C for the Co₅₀Gd_{50 − x}Ni_x amorphous alloys, as shown in Fig. 3(c). For comparison, the relationship between the composition and T_C for the Gd₅₀Co_{50 − y}Ni_y (y = 2 and 5) amorphous alloys is also illustrated in Fig. 3(c). The T_C of the Co₅₀Gd_{50 − x}Ni_x amorphous alloys increases from 267 K for Co₅₀Gd₅₀, to 273 K for Co₅₀Gd₄₉Ni₁, 280 K for Co₅₀Gd₄₈Ni₂, and to 289 K for Co₅₀Gd₄₇Ni₃. In contrast, the T_C of the Gd₅₀Co_{50 − y}Ni_y (y = 2 and 5) amorphous alloys decreases from 267 K for y = 0 to 260 K for y = 2 and to 248 K for y = 5.

It is common that in the Gd–(TM)-based amorphous alloys, there exist at least three kinds of interactions between the matrix elements: the direct 3d–3d electronic interaction between the transition elements that contain the unfilled 3d shell and the indirect coupling of 4f–4f electronic interaction form the RE elements. Besides the respective interactions of the compositional elements themselves, there also exists the 3d–4f interaction between the different atoms. Referring to the previous research on the Gd–TM amorphous alloys,^[14,25] the 3d–4f indirect interaction is considered to be negligible compared with other interactions in determining the Curie temperature of the samples. Therefore, for Gd₅₀Co_{50 − y}Ni_y (x = 2 and 5) amorphous alloys with a constant proportion of RE elements, the changed T_C of the alloys should have little relation with the RE interactions and could be mainly ascribed to the 3d–3d direct interaction between the transition metals.^[20] On the other hand, the decreasing T_C with Ni content in the Gd₅₀Co_{50 − y}Ni_y (y = 2 and 5) amorphous alloys is understandable because the magnetic moment of Ni is smaller than that of Co, which means the reduced 3d–3d interactions between the transition metals. The effect of Ni substitution on the T_C of the Co₅₀Gd_{50 − x}Ni_x metallic glasses, however, is more complicated. The nominal magnetic moment (μ_nom.) of the Co₅₀Gd_{50 − x}Ni_x amorphous alloys decreases from 6.352μ_B for x = 1, to 6.302μ_B for x = 2 and to 6.256μ_B for x = 3 because the magnetic moment of Ni is much smaller than that of Gd. However, the effective magnetic moment (μ_eff) calculated according to the Curie–Weiss law, as shown in the inset of Fig. 3(b), increases from 5.38μ_B for x = 1, to 5.8μ_B for x = 2 and to 7.87μ_B for x = 3. The increasing μ_eff implies that the Ni substitution for Gd enhances the magnetic interactions with the increasing Ni concentration. According to the results obtained from the Gd–Ni binary amorphous alloy system,^[14] the interactions between Gd–Gd and Gd–Ni in the system are similar to each other in determining the T_C. Therefore, for the Co₅₀Gd_{50 − x}Ni_x amorphous alloys with constant Co concentration, the Co–Co interactions are almost unchanged, and thus, the enhanced magnetic interaction and the improved T_C with Ni addition in the Co₅₀Gd_{50 − x}Ni_x amorphous ribbons are probably ascribed to the additional 3d–3d coupling from the Co–Ni interactions.

The temperature dependence curves of −ΔS_m under various magnetic fields of the Co₅₀Gd_{50 − x}Ni_x amorphous alloys were obtained by measuring the magnetic performance as shown in Fig. 4. All the Co₅₀Gd_{50 − x}Ni_x amorphous alloys show a typical λ-shape within a broader temperature range as listed in Table 1. The is increased to 4.9 J · kg⁻¹·K⁻¹ for Co₅₀Gd₄₉Ni₁ compared with that of the Co₅₀Gd₅₀ amorphous alloy, and then decreased with continuous increase of the Ni content from 2 at.% to 3 at.%. This is anomalous because it is generally recognized from the relationship that the in the Gd–TM-based amorphous alloys shows an approximately linear relationship with .^[28] To investigate the relationship between and T_C of the Co₅₀Gd_{50 − x}Ni_x amorphous alloys in more detail, the relationship vs. for each sample is plotted in Fig. 5. Unlike the previous results obtained from the Gd–TM-based metallic glasses, the dependence of on for the Co₅₀Gd_{50 − x}Ni_x (x = 0, 1, 2, and 3) amorphous ribbons is somewhat like a superposition of a linearly decreasing and a parabolic-like relationship. The parabolic-like relationship between and is supposed to be resulted from the additional 3d–3d coupling from the interaction of Co–Ni, as the extra 3d–3d coupling is absent in the Co₅₀Gd₅₀ binary amorphous alloy and achieves the maximum effect in the Co₅₀Gd₄₉Ni₁ amorphous alloy. Consequently, T_C and of the binary Gd₅₀Co₅₀ metallic glass are improved simultaneously through 1 at.% Ni addition.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. The −ΔSm vs. T curves for the Co50Gd50 − xNix amorphous ribbons: (a) x = 1, (b) x = 2, and (c) x = 3.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. The vs. plots for the Co50Gd50 − xNix (x = 1, 2, and 3) amorphous ribbons under various magnetic fields from 1 T to 5 T.

From another point of view, the cooling efficiency, as an important parameter of the magnetic refrigeration material, can be evaluated in a more intuitive way through considering the adiabatic temperature rise (ΔT_ad) of the alloys. By measuring the heat capacity C_p(T) of the sample, as shown in the inset of Fig. 6, the adiabatic temperature rise (ΔT_ad) can be obtained as

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. The ΔTad–T curve of the Co50Gd49Ni1 and Co50Gd47Ni3 amorphous ribbons under the fields of 2 T and 5 T, the inset is the Cp–T curves of the amorphous alloys.

In summary, we added a small amount of Ni to Gd₅₀Co₅₀ binary amorphous alloy and obtained Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) amorphous ribbons with improved Curie temperature. The glass forming ability remains stable under the substitution of 1 at.% Ni, while the substitution of 2–3 at.% Ni weakens the thermal stability and GFA of the alloy. The MCE of the Co₅₀Gd_{50 − x}Ni_x (x = 1, 2, and 3) amorphous alloys was studied. It was found that with the increasing T_C of the Co₅₀Gd_{50 − x}Ni_x amorphous alloys, increases to about 4.9 J · kg⁻¹ · K⁻¹ for x = 1, and then decreases to 4.46 J · kg⁻¹ · K⁻¹ for x = 2 and 4.15 J · kg⁻¹ · K⁻¹ for x = 3. The compositional dependence on the Curie temperature as well as of the Co₅₀Gd_{50 − x}Ni_x metallic glasses, and the mechanism involved, were investigated. It was believed that the improved T_C by minor Ni substitution for Gd, and the deviation of vs. plots of the Co₅₀Gd_{50 − x}Ni_x (x = 0, 1, 2, and 3) amorphous alloys from the generally recognized linear relationship, are most likely due to the additional Co–Ni interaction.