Magnetocaloric response of amorphous and nanocrystalline Cr-containing Vitroperm-type alloys

Magnetocaloric response of amorphous and nanocrystalline Cr-containing Vitroperm-type alloys L.M. Moreno-Ramírez1, J.S. Blázquez1, V. Franco1, A. Conde1, M. Marsilius2, V. Budinsky2, G. Herzer2 1 Dpto. Física de la Materia Condensada, ICMSE-CSIC, Universidad de Sevilla, P.O. Box 1065, 41080Sevilla, Spain 2 Vacuumschmelze GmbH & Co KG, Grüner Weg 37, D-63450 Hanau, Germany Abstract The broad compositional range in which transition metal (TM) based amorphous alloys can be obtained, yields an easily tunable magnetocaloric effect (MCE) in a wide temperature range. In some TM-based alloys, anomalous behaviors are reported, as a non-monotonous trend with magnetic moment (e.g. FeZrB alloys). Moreover, in certain Cr-containing Vitroperm alloys anomalously high values of the magnetic entropy change were published. In this work, a systematic study on MCE response of Crcontaining amorphous alloys of composition Fe74-xCrxCu1Nb3Si15.5B6.5 (with x=2, 8, 10, 12, 13, 14 and 20) has been performed in a broad Curie temperature range from 100 K to 550 K. Curie temperature and magnetic entropy change peak of the amorphous alloys decrease with the increase of Cr content at rates of -25.6 K/at.% Cr and -54 mJkg-1K1/at.% Cr, respectively, following a linear trend with the magnetic moment in both cases. The presence of nanocrystalline phases has been considered as a possible cause in order to explain the anomalies. The samples were nanocrystallized in different stages, however, the magnetocaloric response decreases as crystallization progresses due to the large separation of the Curie temperatures of the two phases. Introduction Magnetic refrigeration is called to be an energetic efficient green alternative to the usual refrigeration systems based on the compression-expansion of gasses. Magnetic refrigeration is based on the magnetocaloric effect (MCE), defined as the temperature change under a magnetic field variation in an adiabatic process (∆Tad) or as the magnetic entropy change in an isothermal process due to a magnetic field variation (∆SM) [1]. From Maxwell relations ∆SM can be expressed as: ∆𝑆𝑀= 𝜇0∫(𝜕𝑀 𝜕𝑇)𝐻 𝐻 0d𝐻, where M, T, H and µ0 are the magnetization, temperature, magnetic field and magnetic permeability of vacuum, respectively. From that, it is deduced that the MCE is maximum around a phase transition which implies magnetization changes, either with first (FOPT) or second order (SOPT) character. FOPT materials show high magnetocaloric peaks but in a narrow temperature range. Moreover, they normally present associated hysteresis phenomena. On the other hand, SOPT materials show smaller magnetocaloric peaks but in a wider temperature range than FOPT materials and without associated hysteresis. In this sense, transition metal (TM) based amorphous alloys have been widely studied since they were proposed as low cost candidates for magnetocaloric applications due to the easily tuned Curie temperature (TC), negligible magnetic hysteresis, high electrical resistivity (decreasing eddy current losses) and excellent mechanical properties [2-5]. In addition to this potential for applications, they constitute model systems in which detailed studies of the different magnitudes affecting magnetocaloric response could be analyzed, like scaling of the properties [6], influence of demagnetizing field [7] or interphase interactions [8], etc. The MCE response of Cr-containing amorphous alloys has been reported with extremely high ∆SMpeak values [9]. However, these values are not in agreement with other data found in the literature [10-12]. These anomalies could be due to a nonmonotonous behavior of the magnetic moment with the addition of different elements in the amorphous matrix (as it has been found for FeZrB alloys [13]) or to a composite character of the system [14]. In this work, a systematic study on the MCE response of a Cr-containing amorphous alloy series has been performed. This series extends in a very broad Curie temperature range, from ~ 100K to 550 K. The possible causes of the described anomalies are explored. Experimental Ribbon samples of composition Fe74-xCrxCu1Nb3Si15.5B6.5 (with x=2, 8, 10, 12, 13, 14 and 20) were produced by melt spinning. Throughout the paper, samples are referenced by its Cr content. Microstructural characterization of the samples was done by x-ray diffraction (XRD) experiments with Cu Kα radiation in a Bruker D8I diffractometer. The study of the thermal stability and the heat treatments of the samples was performed by differential scanning calorimetry (DSC) in a Perkin–Elmer DSC7 under Ar flow. Magnetic measurements were performed in a Lakeshore 7407 vibrating sample magnetometer (VSM) equipped with a cryostat down to 77.7 K or an oven up to 1300 K. Samples with ~25 𝜇m of thickness were cut as discs of 3 mm diameter to minimize the effect of the demagnetizing factor and sample positioning. The mass of the samples was measured in a Mettler Toledo XP26 scale with an accuracy of 1 𝜇g. Magnetocaloric analysis has been performed using the software Magnetocaloric Effect Analysis Program [15], available from LakeShore Cryotronics Inc. Results Figure 1 shows the XRD patterns of the as-quenched Cr-containing Vitroperm ribbons in which it can be observed that samples are fully amorphous, independently of the Cr content. Concerning the thermal stability, figure 2 shows the DSC scans of the asquenched amorphous ribbons. From these scans up to 993 K at 20 K/min, two transformation processes can be observed, although the high temperature one shifts out of the explored range as the Cr content decreases. The first process corresponds to the crystallization of the bcc-Fe nanophase (α-Fe), as confirmed by XRD experiments (figure 7). The increase of Cr content shifts the nanocrystallization peak to higher temperatures, stabilizing the amorphous structure. For the sample with 20 at.% Cr, the primary crystallization peak overlaps with high temperature processes and the nanocrystalline state is not observed. Magnetic entropy change curves were calculated using the software previously mentioned using a numerical approximation to the Maxwell equation. Figure 3 shows the specific magnetization, σ(T), at 200 Oe (upper panel) and ∆SM(T) for a magnetic field change of 15 kOe (lower panel) for the amorphous ribbons. From σ(T) data it can be observed that the Curie temperature of the amorphous alloys decreases as Cr content increases. This variation follows a linear dependence with a slope of -25.6±0.8 K/at.% of Cr, in agreement with previous studies [16]. Although the increase in Cr content allows us to reduce TC, it is at the expense of a decrease of the magnetic entropy change peak (∆SMpeak) (figure 4). The refrigerant capacity (RC), calculated as the product of ∆SMpeak times the full width half maximum (FWHM), also decreases with the increase of Cr content, although the variation is less significant than for ∆SMpeak (comparing samples with 2 and 14 at.% of Cr, the variation of the ∆SMpeak is about 56%, meanwhile, for RC it is 18%). For the sample with 20 at.% Cr, RC cannot be calculated because the FWHM is out of the cryostat range. Both variations of ∆SMpeak and RC are explained by the reduction of magnetic moment with the increase of Cr content in Vitroperm alloys following a dilution law [17,18], as can be observed in inset of figure 5. In fact, a linear evolution between ∆SMpeak and the magnetic moment in Fe atoms (µFe) can be observed (figure 5). The magnetocaloric response of the studied amorphous ribbons have been compared to other previously studied amorphous series including bulk amorphous [1921], Vitroperm [10-12,22,23] and Nanoperm [2,24-45] compositions, as it is shown in figure 6. Values from literature were rescaled to 10 kOe using a power law with n=0.75 [46]. The values obtained in this study for the samples containing Cr are in agreement with those Vitroperm type alloys without anomalous response [10-12]. Although the anomalous high values were assigned for amorphous samples, in order to explore a suitable cause for anomalies in MCE response the presence of small amount of nanocrystals have been considered. The samples were nanocrystallized to produce composite materials. The thermal treatments were performed in the differential scanning calorimeter at two stages of the first transition process, T1 and T2, corresponding to the mid-point temperature between the onset and the peak temperatures and to the peak temperature, respectively (figure 2). With respect to the microstructural characterization, crystal size (D) and crystalline fraction (XC) were obtained from XRD experiments (figure 7). D was estimated using the Scherrer formula and XC was obtained from the area ratio between the amorphous halo and the (110) α-Fe line. The deconvolution of amorphous halo and crystalline peak has been done using Gaussian and Lorentzian peaks, respectively (figure 8). For both heat treatments, the crystalline fraction is reduced and the crystal size increases as Cr content increases in the alloy. Combining these two facts it is deduced that Cr hinders the nucleation of the α-Fe nanocrystals in the amorphous matrix. With respect to the magnetocaloric response, figure 9 shows ∆SM for samples nanocrystallized with 2 and 14 at.% Cr along with the response of the corresponding amorphous alloy for comparison. The magnetocaloric response clearly decreases as the crystallization fraction increases which is due to the high Curie temperature of the crystalline phase with respect to that of the amorphous phase, reducing the amount of phase that has a significant MCE response at the studied temperature. Therefore nanocrystallization cannot explain the extremely high ΔSM values reported earlier for Cr-containg alloys (6). Thus these previous results appear to be due experimental artifacts still to be clarified. It can also be observed that crystallization of α-Fe phase changes the composition of the residual amorphous matrix, varying its Curie temperature. This fact explains why the high temperature response of the samples is not monotonous with the crystalline fraction. Changing the temperature axis (used in figure 9) to a reduced temperature axis, t=(T-TC)/TC, the expected monotonous behavior with the increase in crystalline fraction is recovered (insets of figure 9). The magnetic entropy change can be expressed as a power law of the form ∆SM=aHn, where n is field independent for single phase materials in three regimes: well below TC (n=1), well above TC (n=2) and at TC (related with the critical exponents of the transition) [46]. TM-based amorphous alloys shows similar values of exponent n at TC around 0.75, as it is observed in our amorphous samples. Figure 10 shows, as an example, n(T) curves for the amorphous and nanocrystalline samples with 2 at.% Cr. The influence of the crystalline phase (ferromagnetic α-Fe impurities) in the nanocrystallized samples can be observed in the departure of the exponent n at TC and above TC from the predicted values for single phase materials. Inset of figure 10 shows the evolution of the exponent n(TC) with the crystalline fraction obtained from XRD experiments, in which n departs to higher values than that typically found in amorphous alloys (~0.75). Above TC, the influence of the ferromagnetic α-Fe (n=1) phase does not allow to achieve the expected value for the paramagnetic phase (n=2). Conclusions In this work, the influence of the Cr content on the magnetocaloric response of amorphous Vitroperm-type ribbons is studied. The increase in Cr content reduces TC at a rate of -25.6±0.8 K/at.% Cr, as well as the magnetic entropy change peak, at a rate of -54 mJkg-1K-1/at.% Cr. This reduction of the magetocaloric response is related to the decrease of the magnetic moment as Cr content increases. The obtained values are in agreement with the general trend observed for TM-based amorphous alloys and a reduction of TC is observed at the expense of a decrease of the MCE response. It has been found that the nanocrystallization of the samples deteriorates the magnetocaloric response due to the large difference between the Curie temperatures of the crystalline and amorphous phases. Acknowledgements This work was supported by the Spanish MINECO (project MAT2013-45165-P) and the PAI of the Regional Government of Andalucía. L.M. Moreno-Ramírez acknowledges a FPU fellowship from the Spanish MECD. 20 40 60 80 100 120 2 at.% Cr 8 at.% Cr 10 at.% Cr 12 at.% Cr 13 at.% Cr 14 at.% Cr 2 (deg) 20 at.% Cr Intensity (a.u.) Figure 1. XRD patterns of the as-quenched Cr-containing Vitroperm alloys. 700 800 900 1000 2 at.% Cr 2.5 W/g T1 8 at.% Cr T2 10 at.% Cr 12 at.% Cr 13 at.% Cr 14 at.% Cr Heat Flow T (K) 20 at.% Cr Figure 2. DSC scans of the as-quenched Cr-containing Vitroperm alloys. 0 20 40 60 100 200 300 400 500 600 700 -1.5 -1.0 -0.5 0.0 2 at.% Cr 13 at.% Cr 8 at.% Cr 14 at.% Cr 10 at.% Cr 20 at.% Cr 12 at.% Cr  (emu/g) H=200 Oe H=15 KOe SM (J K-1 kg-1) T (K) Figure 3. Specific magnetization (σ) at 200 Oe (upper panel) and magnetic entropy change (∆SM) up to 15 kOe (lower panel) for the amorphous ribbons. 100 200 300 400 500 600 0.5 1.0 1.5 0 5 10 15 20 80 85 90 95 Tpeak (K) H=15 kOe slope=-25.6±0.8 K/at.% Cr slope=-0.054±0.007 (Jkg-1K-1)/at.% Cr RC (J kg-1)|Speak M| (J K-1 kg-1) Cr content (at.%)