Excellent magnetocaloric properties in RE₂Cu₂Cd (RE = Dy and Tm) compounds and its composite materials

Abstract

The magnetic properties and magnetocaloric effect (MCE) of ternary intermetallic RE₂Cu₂Cd (RE = Dy and Tm) compounds and its composite materials have been investigated in detail. Both compounds undergo a paramagnetic to ferromagnetic transition at its own Curie temperatures of T_C ~ 48.5 and 15 K for Dy₂Cu₂Cd and Tm₂Cu₂Cd, respectively, giving rise to the large reversible MCE. An additionally magnetic transition can be observed around 16 K for Dy₂Cu₂Cd compound. The maximum values of magnetic entropy change (−ΔS_M^max) are estimated to be 17.0 and 20.8 J/kg K for Dy₂Cu₂Cd and Tm₂Cu₂Cd, for a magnetic field change of 0–70 kOe, respectively. A table-like MCE in a wide temperature range of 10–70 K and enhanced refrigerant capacity (RC) are achieved in the Dy₂Cu₂Cd - Tm₂Cu₂Cd composite materials. For a magnetic field change of 0–50 kOe, the maximum improvements of RC reach 32% and 153%, in comparison with that of individual compound Dy₂Cu₂Cd and Tm₂Cu₂Cd. The excellent MCE properties suggest the RE₂Cu₂Cd (RE = Dy and Tm) and its composite materials could be expected to have effective applications for low temperature magnetic refrigeration.

Introduction

Magnetic refrigeration technology based on the magnetocaloric effect (MCE) shows superior application potential over conventional gas compression/expansion refrigeration technology because of its environmental friendliness, higher energy efficiency as well as compactness^1,2,3,4,5. The MCE is an intrinsic thermal response for the application or removal of a magnetic field to a magnetic material, which can be characterized by the coupled variations of two quantities: the adiabatic temperature change (ΔT_ad) or/and isothermal magnetic entropy change (ΔS_M). To satisfy practical application, extensive efforts have been carried out to pick out the magnetic materials with large/giant MCE as magnetic refrigerants^{1,2,3,4,5,6,7,8,9,10}.

Recently, the rare-earth (RE) based alloys and oxides, which exhibit the large reversible MCEs and refrigeration capacity with small or zero hysteresis have been of great of interest^{11,12,13,14,15,16,17}. Increasing efforts have been devoted for study of the ternary intermetallic compounds of the RE₂T₂X2:2:1 (T = transition metals and X = III group p-metals). Among of the 2:2:1 system, the RE₂Cu₂X (X = Mg, Cd, Sn or In) crystallized with the tetragonal Mo₂B₂Fe-type structure¹⁸, have attracted some attentions because of their unique physical and magnetic properties. The basic crystal chemical data of the different RE₂T₂X series have been reviewed^19,20. Very recently, Zhang et al. and Li et al. have reported the large reversible MCEs in RE₂Cu₂In (RE = Dy, Er and Tm) and Ho₂T₂In (T = Cu and Au) compounds, respectively^21,22,23. However, the systems with the p-metals as cadmium are much less known, what besides other reasons could be explained also by the difficulty in materials synthesis due to the high vapour pressure (low boiling point) of cadmium.

To further understand the physical properties of RE₂T₂X system, in this paper, the magnetic properties and MCE in RE₂Cu₂Cd (RE = Dy and Tm) compounds and its composite materials have been investigated systematically. Not only a large reversible MCE was observed in Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds, but also an enhanced refrigerant capacity was found in its composite materials.

Results and Discussion

Figure 1(a,b) show the temperature dependence of the zero field cooled (ZFC) and field cooled (FC) magnetization M under different magnetic fields for Dy₂Cu₂Cd and a magnetic field of 2 kOe for Tm₂Cu₂Cd, respectively. Both compounds display a typical paramagnetic to ferromagnetic (PM-FM) transition and the Curie temperatures T_C, corresponding to the peak of dM_FC/dT - T curve [inset of Fig. 1(b)], are determined to be 48.5 K and 15 K for Dy₂Cu₂Cd and Tm₂Cu₂Cd, respectively. Another magnetic transition can be observed for Dy₂Cu₂Cd around T_S ~ 16 K under low magnetic fields and it shifts to much lower temperatures with increasing magnetic field. Such behaviours may arise from a spin glass transition or spin reorientation phenomenon^24,25, a systematically detail study of the lower temperature magnetic transition will be performed later. The transition temperatures are in good agreement with previously reported values in the literatures²⁰. Figure 2(a,b) show the temperature dependence of the magnetization M (left side) and the reciprocal susceptibility 1/χ (right side) for Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds under a magnetic field of 10 kOe, respectively. The 1/χ in paramagnetic regime from 80 to 298 K obeys the Curie-Weiss law for both compounds. The fitted lines are a guide to the eyes for Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds as shown in the insets of Fig. 2(a,b), respectively. The fit to the Curie-Weiss formula yields positive paramagnetic Curie temperatures (θ_P), θ_P = 45.3 K for Dy₂Cu₂Cd and θ_P = 14.1 K for Tm₂Cu₂Cd, respectively, suggesting dominant ferromagnetic interactions. The effective magnetic moments (μ_eff) are 10.84 μ_B and 7.72 μ_B for Dy₂Cu₂Cd and Tm₂Cu₂Cd, respectively. Such moments are close to those of the free ion values of Dy and Tm taking the theoretical RE³⁺ moment of 10.86 μ_B and 7.56 μ_B, respectively.

Figure 1

Temperature dependence zero-field cooling (ZFC) and field cooling (FC) magnetization (M) under different magnetic fields for Dy₂Cu₂Cd (a) and the magnetic field of 2 kOe for Tm₂Cu₂Cd (b) compounds, respectively. Inset of (b) shows the temperature dependence dM_FC/dT for Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds under the magnetic field of 2 kOe.

Figure 2

Temperature dependence of magnetization (M, left scale) and the reciprocal susceptibility (1/χ, right scale) for Dy₂Cu₂Cd (a) and Tm₂Cu₂Cd (b) compounds under the magnetic field of H = 10 kOe, respectively.

The magnetic isothermal M(H) curves of Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds with increasing field around their transition temperatures with increasing magnetic field up to 70 kOe have been measured and some of them are shown in Figs 3(a) and 4(a), respectively. The magnetization below T_C increases rapidly in the low magnetic field range for both compounds and it tends to saturate for Dy₂Cu₂Cd compound with increasing magnetic field, whereas it is not saturated at 70 kOe for Tm₂CuCd compound. To further understand the magnetic transitions, Arrott plots (H/M vs. M²) of Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds are shown in Figs 3(b) and 4(b), respectively. According to Banerjee criterion²⁶, the signal (positive and negative) of the slope in Arrott plots has been used to determine the nature of the magnetic phase transition. The negative slopes or inflection points in the Arrott plots often are corresponding to a first order phase transition, whereas the positive slopes are associated to a second order phase transition. By this criterion, neither the inflection points nor negative slopes can be observed in the Arrott plots for Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds, indicating a characteristic of the second order (FM-PM) magnetic phase transition.

Figure 3

(a) Magnetic field dependence of the magnetization (increasing field only) for Dy₂Cu₂Cd at some selected temperatures. (b) The plots of H/M versus M² for Dy₂Cu₂Cd at some selected temperatures.

Figure 4

(a) Magnetic field dependence of the magnetization (increasing field only) for Tm₂Cu₂Cd at some selected temperatures. (b) The plots of H/M versus M² for Tm₂Cu₂Cd at some selected temperatures.

Figure 5(a,b) show the temperature dependence of magnetic entropy change −ΔS_M for Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds which is derived from the temperature and field dependence of the magnetization M (H, T) by using the Maxwell’s thermodynamic relation²⁷, , respectively. It can be found that the maximum value of −ΔS_M increases monotonically with increasing magnetic field change for both compounds [see insets of Fig. 5(a,b)]. Two successive −ΔS_M peaks (one at around T_C, another at around T_S) can be clearly seen even the low magnetic field change for Dy₂Cu₂Cd compound, thus obviously enlarging the temperature range of MCE. Only a pronounced peak in the −ΔS_M(T) curves is observed around T_C for Tm₂Cu₂Cd compound. For the magnetic field changes of 0–20, 0–50 and 0–70 kOe, the maximum values of the magnetic entropy change (−ΔS_M^max) are evaluated to be 7.2, 13.8 and 17.0 J/kg K around T_C and 3.3, 6.6 and 8.3 J/kg K around T_S for Dy₂Cu₂Cd compound; and to be 9.2, 17.3 and 20.8 J/kg K for Tm₂Cu₂Cd compound, respectively.

Figure 5

The magnetic entropy change −ΔS_M as a function of temperature for various magnetic field changes ΔH up to 0–70 kOe for Dy₂Cu₂Cd (a) and Tm₂Cu₂Cd (b) compounds, respectively. Insets of (a,b) show the maximum values of magnetic entropy change (−ΔS_M^max) as a function of the magnetic field changes for Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds, respectively.

In addition, the ΔS_M (T) curves for the materials with the second order phase transition can be also described using a universal curve^28,29, which is constructed by normalizing with their respective maximum value ΔS_M^max (i. e. ΔS′ = ΔS_M (T)/ΔS_M^max) and rescaling the temperature θ, defined as

where the T_r1 and T_r2 are the temperatures of the two reference points of each curve that correspond to 0.6ΔS_M^max. The transformed ΔS′ (θ) curves for Tm₂Cu₂Cd and Dy₂Cu₂Cd compounds are displayed in Figs 6 and 7, respectively. We can note that all the rescaled ΔS_M curves for Tm₂Cu₂Cd are overlapped with each other in the present temperature range, as shown in Fig. 6, proving the occurrence of the second order magnetic phase transition in Tm₂Cu₂Cd compound. In parallel, the curves for Dy₂Cu₂Cd compound are also overlapped with each other around and above T_C (see Fig. 7). Whereas an obvious deviation below T_C for θ < −2 (around T_S) can be found which is properly due to the spin reorientation phenomenon or spin glass transition. Therefore, the ΔS_M (T) around T_S (5–30 K) are rescaled and the results are shown in the inset of Fig. 7. Similarly, the curves around T_S are well overlapped with each other. Furthermore, the rescaled ΔS′ (θ) curves around T_C and T_S for Dy₂Cu₂Cd compound under various magnetic field changes are summarized together (as given in the Fig. 8). One can find that all the rescaled ΔS_M curves can collapse onto one universal curve, which is consistent with the previous investigations that the materials with successive magnetic phase transitions^22,24,30,31. The analysis of the universal behaviour further confirms that the Dy₂Cu₂Cd compound with the second order phase transition.

Figure 6

Normalized magnetic entropy change ΔS′ (=ΔS_M/ΔS_M^max) as a function of the rescaled temperature θ in the present temperature range for Tm₂Cu₂Cd compound.

Figure 7

Normalized magnetic entropy change ΔS′ (=ΔS_M/ΔS_M^max) as a function of the rescaled temperature θ around T_C for Dy₂Cu₂Cd compound.

Inset shows the normalized magnetic entropy change ΔS′ (=ΔS_M/ΔS_M^max) as a function of the rescaled temperature θ around T_S for Dy₂Cu₂Cd compound.

Figure 8

Normalized magnetic entropy change ΔS′ (=ΔS_M/ΔS_M^max) as a function of the rescaled temperature θ around T_C and T_S for Dy₂Cu₂Cd compound.

Another important quality factor of refrigerant materials is the refrigerant capacity [RC, defined as numerically integrating the area under the −ΔS_M - T curve at full width of half maximum (δ_FWHM) of the −ΔS_M peak as the integrating limits]. For the magnetic field changes of 0–20, 0–50 and 0–70 kOe, the values of RC are evaluated to be 87, 316 and 495 J/kg for Dy₂Cu₂Cd compound; and to be 60, 165 and 248 J/kg for Tm₂Cu₂Cd compound, respectively. It is well known that magnetic refrigeration systems based on an ideal Ericsson cycle requires a magnetocaloric material with a constant ΔS_M over an operating refrigeration temperature range^32,33. Besides the materials with successive magnetic transitions or with a very magnetic field sensitive magnetic phase transitions^{22,24,30,31,34}, composite materials have been considered to be the most promising method to accomplish the requirement of Ericsson cycle since it can lead to almost constant ΔS_M with enlarged temperature span^35,36,37,38. An enhanced RC have been successfully realized in Eu₈Ga₁₆Ge₃₀-EuO³⁶, amorphous FeZrB(Cu)³⁷ and ErNiBC-GdNiBC³⁸ composite materials. We can note that the Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds possess the same crystal structure with similar lattice parameters and similar magnitudes of the magnetic entropy change (−ΔS_M). Therefore, these composite materials could be expected to fulfil the required Ericsson cycle conditions. The total magnetic entropy change of x Dy₂Cu₂Cd + (1 − x) Tm₂Cu₂Cd composite materials, ΔS_comp(T, H, x), can be calculated theoretically from the individual ΔS_M(T) curves^35,36,37,38,

where x and 1 − x are the weight amounts of Dy₂Cu₂Cd and Tm₂Cu₂Cd, respectively. Based on both compounds, a composite material can be formed and the optimum ratio of x ~ 0.77 is determined by using a numerical method. The magnetic entropy change ΔS_comp(T) for Dy₂Cu₂Cd - Tm₂Cu₂Cd composite material at x ~ 0.77 under a magnetic field change of 0–50 kOe is shown in Fig. 9. A table-like MCE in a wide temperature span of 10–70 K can be observed in ΔS_comp(T) curve which is desirable for an ideal Ericsson-cycle magnetic refrigeration. The corresponding maximum value of RC_comp is 417 J/kg, which is 32% and 153% higher than those of Dy₂Cu₂Cd (316 J/kg) or Tm₂Cu₂Cd (165 J/kg). The transition temperature T_C, the maximum values of −ΔS_M^max and RC under the magnetic field change of 0–50 kOe for Dy₂Cu₂Cd and Tm₂Cu₂Cd as well as the 0.77 Dy₂Cu₂Cd - 0.23 Tm₂Cu₂Cd composite material together with some MCE materials in the similar working temperature range are listed in Table 1 for comparison. The MCE parameters for the present studied materials are comparable or larger than those of other potential magnetic refrigerant materials in the similar temperature region, suggesting RE₂Cu₂Cd composite materials could be a promising candidate for magnetic refrigeration for Ericsson cycle in the temperature range of 10–70 K. The present results allow for the possibility of using RE₂Cu₂Cd compounds to fabricate composite materials with desirable magnetocaloric properties for active magnetic refrigeration.

Table 1 The transition temperature T_C, the maximum values of magnetic entropy change −ΔS_M^max and refrigeration capacity RC under the magnetic field change of 0–50 kOe for Dy₂Cu₂Cd and Tm₂Cu₂Cd as well as the 0.77Dy₂Cu₂Cd - 0.23Tm₂Cu₂Cd composite material together with some MCE materials with the T_C from 10 to 70 K.

Figure 9

Temperature dependence of magnetic entropy change −ΔS_comp for the 0.77 Dy₂Cu₂Cd - 0.23 Tm₂Cu₂Cd composite material for the magnetic field change of 0–50 kOe.

Conclusions

In summary, two single phased Dy₂Cu₂Cd and Tm₂Cu₂Cd compounds have been fabricated and the magnetism and magnetocaloric effect have been investigated experimentally. Both compounds undergo a paramagnetic to ferromagnetic transition at their own Curie temperatures, additionally, another magnetic transition is also observed for Dy₂Cu₂Cd at low temperatures. For a magnetic field change of 0–50 kOe, the maximum values of magnetic entropy change (−ΔS_M^max) are 13.8 J/kg K around T_C and 6.6 J/kg K around T_S for Dy₂Cu₂Cd; and 17.3 J/kg K for Tm₂Cu₂Cd, respectively. The rescaled entropy change ΔS_M curves around T_C follow a universal behaviour for Dy₂Cu₂Cd and Tm₂Cu₂Cd, which further confirm both compounds with the second order phase transition. A table-like MCE from 10 to 70 K and a strong enhancement of RC have been found in theoretically calculated Dy₂Cu₂Cd-Tm₂Cu₂Cd composite materials. The maximum value of RC_comp is 417 J/kg in the 0.77Er₂Cu₂Cd - 0.23Tm₂Cu₂Cd composite material for a magnetic field change of 0–50 kOe, which is obviously larger than those for either Dy₂Cu₂Cd (316 J/kg) or Tm₂Cu₂Cd (165 J/kg) compounds. The results indicate that the RE₂Cu₂Cd (RE = Dy and Tm) compounds and its composite materials could be promising candidates for magnetic refrigeration in the temperature range of 10–70 K. Furthermore, the present results may also provide a cost-effective strategy for exploring suitable refrigeration candidates with table-like magnetocaloric feature by a materials composition method, beneficial for Ericsson-cycle in the wide temperature range.

Methods

The Dy₂Cu₂Cd and Tm₂Cu₂Cd polycrystallize samples were fabricated by induction melting the elements in a sealed quartz crucible. Firstly, high purity Dy, Tm, Cu and Cd with stoichiometric amounts were weighted and placed in the quartz crucible. Secondly, a high vacuum better than 2*10⁻⁵ mbar was achieved in the crucible. Then the crucible was filled with purified argon gas at pressure of ca. 750 mbar and sealed immediately. Finally, the quartz crucible was placed in an induction furnace and heated at 1100 K for 4 minutes, following by 3 hours annealing at 850 K. The powder X-ray diffraction (Bruker D8 Advance) measurements were carried out at room temperature using Cu Kα radiation. Both samples were proved to be single phase and the lattice parameters were evaluated to be a = 7.491 and c = 3.742 Å for Dy₂Cu₂Cd; and to be a = 7.439 and c = 3.687 Å for Tm₂Cu₂Cd, respectively. The magnetic measurements were performed by using a commercial vibrating sample magnetometer (VSM) which is an option of the physical properties measurement system (PPMS-9, Quantum Design) in the temperature range of 3–298 K with a DC magnetic field from 0 to 7 T and the samples are small particles of 4.5 and 3.8 mg for Dy₂Cu₂Cd and Tm₂Cu₂Cd, respectively.