Magnetocaloric properties of Gd(Co1-xFex)2 compounds, with x ≤ 0.60

In this paper the results of specific magnetization and magnetocaloric effect (MCE) measurements for Gd(Co1-xFex)2 system upon the Co substitution by Fe for the x = 0 ÷ 0.60 range are presented. Phase composition was controlled by X-ray diffraction analysis. MCE has been studied within the temperature range of 300-850 K in magnetic fields up to 17 kOe by the magnetic entropy change calculation (∆Sm). It was found that in contrast to the previously studied R(Co-Fe)2 compounds where R = Dy, Ho, Er, an ordinary symmetrical peak of ΔSm(T) in the vicinity of TC is observed for presented samples. Additionally, the MCE comparison of Gd(Co0.88Fe0.12)2 with that for the isostructural Gd(Ni0.88Fe0.12)2 compound having a plateau-like ∆Sm temperature dependence is given. The obtained results are discussed.


Introduction
In the last two decades intensive research has been devoted to the search for and development of new magnetocaloric materials for the purpose of their application in energy-efficient and environment-friendly systems of magnetic cooling. Different types of compounds with large magnetocaloric effect (MCE) in a narrow temperature interval around the magnetic phase transition temperature were pointed out: RCo2 (Rheavy rare-earth metal), Gd5Ge2Si2, MnFe(P1-xAsx), La(Fe13-xSix), Ni-Mn-Ga, et al. [1]. Nevertheless, the study of the magnetic properties and the MCE in the R(Me1-xFex)2 (R = Tb, Ho; Me = Al, Ni, Co) quasi-binary systems shown that a partial substitution of Me-element by Fe leads to Curie temperature (TC) increase, and the emergence of significant MCE in a wide temperature range below its TC [2][3][4][5]. The latter is an important material property when it is used for the manufacture of magnetic refrigerator working bodies.
Our recent MCE measurements for some Dy(Co1-xFex)2 [6], Ho(Co1-xFex)2, Er(Co1-хFeх)2 [7], and Gd(Ni1-xFex)2 [8] compounds with Ni and Co substitution by Fe confirmed that results and allowed to suggest the reasons of MCE peak widening in the temperature range below their Тc. In general, the results obtained demonstrate a qualitatively identical picture, as in compounds of the R(Co-Fe)2 system with R = Tb, Dy, Ho, Er, and R(Ni-Fe)2 compounds with R = Gd, Tb, Ho . Thus it allows us to lean in favor of the "weak magnetic sublattice" MCE model [9], giving a qualitative interpretation of the magnetocaloric phenomena in the R(Me-Fe)2 intermetallic compounds with Laves phases structure.
These data prompted us to undertake the magnetic and magnetocaloric properties studies of similar to mentioned -Gd(Co1-хFeх)2 system with Fe content (x) from 0 to 0.60. These compounds, unlike those described above with iron concentration absence do not have a Curie point close to room temperature. However, they are of scientific interest from the point of view of the influence of cobalt replacement by iron on the Curie temperature, the magnitude and MCE peak width and refrigerating capacity (q) of these compounds.

Experimental details
Gd(Co1-хFeх)2 alloys were melted in an electric arc furnace under a pure helium protection. An excess of rare earths metal (~ 3 wt. %) was added to the starting compositions the formation preventing of Co-rich phases. A homogenizing annealing of alloys was made in a vacuum furnace at 800 K during 24 hours. The samples structure was determined by X-ray diffraction technique (D8 Advance, Bruker) with Cu Kα radiation source. Diffraction patterns were analyzed by Rietveld method using the "Fullprof" software [10].
Magnetic susceptibility temperature dependence measurements of the samples were carried out in an alternating magnetic field of 50 Oe in the temperature range from 25 to 800° C.
Their magnetization was measured using vibrating sample magnetometer (7407, Lake Shore Cryotronics) in the temperature range from 300 K to 850 K under a magnetic field up to 17 kOe.
Analysis of the X-ray diffraction data at room temperatures showed that all samples contain mainly the 1:2 type Laves phase. Their crystal lattices belong to the Fd-3m space group. For example, figure 1 shows the X-ray diffraction pattern at room temperatures for GdCo2. Fig. 1. The X-ray diffraction pattern of GdCo2.
Also, samples were confirmed to contain only 1:2-phase by measurements of susceptibility temperature dependences (χ(T)) at external magnetic fields of H = 50 Oe. Curie temperatures (TC) were determined from the positions of the dM/dT peaks on the magnetization (M) temperature axis, taken from the specific magnetization temperature dependencies (M(T)) in the magnetic field of 900 Oe (see Fig. 2).  Figure 3 shows the field dependences of the specific magnetization at different temperatures for the Gd(Co0.60Fe0.40)2 compound. Similar isotherms are observed for the samples of other compounds. It can be seen that even at T = 485 K, that is only of 225 K below TC, the magnetization curve practically reaches the saturation. This indirectly indicates that in this temperature range the compound has a low magnetocrystalline anisotropy (MCA) energy and the relatively high degree of magnetic order is realized in Gd-sublattice. Consequently, the probability of the MCE appearance caused by the suppression of Gd atomic magnetic moments disorientation by an external magnetic field is small. For the calculation of isothermal magnetic entropy change (∆Sm), from the field magnetization M(H) isotherms (Fig. 3), we used the following equation [11]: (2) Fig. 4. Temperature dependences of magnetic entropy change ∆Sm(T) in magnetic fields (0-17) kOe for the studied Gd(Co1-xFex)2 compounds. Figure 4 shows the calculated values of ΔSm when the change of magnetic field ∆Н = 17 kOe. It can be seen that, in contrast to the previously studied R(Co-Fe)2 compounds, where R = Dy, Ho, Er [6,7], for the samples with R = Gd, an ordinary symmetrical peak of ΔSm(T) in TC neighborhoods is observed.
Using the ΔSm(T) data, we calculated the refrigerant capacity (q) [11]: This parameter is used for comparison of the different magnetocaloric materials from the practical applications point of view. Values of q are presented in Table 1. Also the values of ∆TFWHM, characterizing the distance between the highest and the lowest temperatures at the half maximum of the -∆S(T) peak, are presented in this Table as well. Consideration of the studied Gd(Co1-xFex)2 compounds under the practical perspective does not cause much enthusiasm: they have small values of ∆S-effect -1 vs 5 J/kgK for Gd (at 17 and 20 kOe, respectively), and hardly can be considered as candidate materials for the manufacture of the working bodies in the magnetic refrigeration systems. Figure 5 presents the relative ∆Sm temperature dependence of Gd(Me0.88Fe0.12)2 (Me = Co, Ni) samples, where the ∆TFWHM value characterizes the MCE peak width [1]. The grey and shaded areas (Fig. 5) correspond to the region of integration at q calculation. These data are 220 J/kg and 41 J/kg at ∆H = 15 kOe for compounds with Ni and Co respectively. It can be seen from the figure that the compound with Ni has a plate-like ΔSm(T) dependence while the compound with Co has a MCE peak in the vicinity of TC typical for many ferromagnets. Hence this large difference in the MCE temperature dependences for the Gd(Me0.88Fe0.12)2 compounds with different Me is a consequence of a strong exchange interactions change in these compounds when Ni is replaced by Co. Figure 6 shows the relative temperature dependences of the parapossess susceptibility -χ(T) of the studied compounds. The χ values were calculated from the linear parts of the M(H) curves at the 10÷15 kOe magnetic fields range (χ = dM/dH). It can be seen that the χ (T) dependence for the compound with Ni in the below TC region lies significantly higher than for the compound with Co (the difference between the curves is shaded); For the sample with Co the χ(T) growth with T rise is much faster.. These differences in the χ(T) variation is an indirect confirmation of the much higher disorder degree in Gd -sublattice for the samples with Ni due to the weaker exchange interaction between Gd and 3d ions sublattices.

Conclusions
Thus, the magnetic and magnetocaloric properties study of the Gd(Co1-xFex)2 (x = 0, 0.08, 0.12, 0.40, 0.60) compounds revealed the following: In these compounds no broadening of the MCE peak observed contrary to that for such compounds with heavier R = Tb, Dy, Ho, Er [5,6]; In contrast to the isostructural Gd(Ni-Fe)2 intermetallics, where the magnetic entropy -ΔSm(T) and paraprocess susceptibility -χ(T) maxima broadenings are observed, in the given compounds, apparently much stronger internal and inter-sublattice exchange interactions keep the magnetic order in both sublattices in the higher degree almost for the whole 0 K ÷ TC temperature interval. So such narrow peaks on ΔSm(T) and χ(T) dependences appear only in the vicinity of TC point.
The obtained results are important for understanding the nature of magnetic and magnetocaloric properties of the large R(Co1-xFex)2 Laves phases group as a whole.