Magnetocaloric effects and electrical resistivity of Ni2Mn0.55CoxCr0.45-xGa – A Heusler alloy system exhibiting a partially-decoupled first-order phase transition

The phase transitions and associated magnetocaloric properties of the Ni2Mn0.55CoxCr0.45-xGa (0 ≤ x ≤ 0.25) Heusler alloy system have been investigated. All samples exhibit a first-order martensitic phase transition, evidenced by a sharp drop in the resistivity versus temperature data and a thermomagnetic irreversibility in the dc magnetization data of the respective samples. Large magnetic entropy changes have also been observed near the phase transitions. The martensitic transformation temperature increases as Cr is partially replaced with Co. Additionally, this substitution leads to a partial decoupling of the magnetic and structural phase transitions, dramatically suppressing any magnetic hysteresis losses. Furthermore, the change in electrical resistivity during the phase transition remains relatively constant across the system, despite major changes in the degree of structural disorder and magnetostructural phase transition coupling. Detailed experimental results and conjectures as to the origin of these behaviors have been provided.The phase transitions and associated magnetocaloric properties of the Ni2Mn0.55CoxCr0.45-xGa (0 ≤ x ≤ 0.25) Heusler alloy system have been investigated. All samples exhibit a first-order martensitic phase transition, evidenced by a sharp drop in the resistivity versus temperature data and a thermomagnetic irreversibility in the dc magnetization data of the respective samples. Large magnetic entropy changes have also been observed near the phase transitions. The martensitic transformation temperature increases as Cr is partially replaced with Co. Additionally, this substitution leads to a partial decoupling of the magnetic and structural phase transitions, dramatically suppressing any magnetic hysteresis losses. Furthermore, the change in electrical resistivity during the phase transition remains relatively constant across the system, despite major changes in the degree of structural disorder and magnetostructural phase transition coupling. Detailed experimental results and conjectures as to the origin of ...


INTRODUCTION
Materials that exhibit coupled structural and magnetic phase transitions are of great importance given that they exhibit several exciting phenomena, [1][2][3] including magnetocaloric effects (MCE). 4he martensitic phase transformation (MPT) is a first-order diffusionless transformation where a high-temperature, high-symmetry crystal structure transforms to a low-temperature, lower-symmetry structure.A selected group of stoichiometric and off-stoichiometric Ni 2 MnX (X = Ga, In, Sb, Sn) Heusler alloy systems have been shown to exhibit a temperature-induced MPT, 5 making these materials attractive from both scientific and application perspectives.][13] The MCE is characterized by the adiabatic temperature change (∆T ad ) or isothermal magnetic entropy change (∆S M ) exhibited by a material as it becomes magnetized. 14Materials (including Heusler alloys) that exhibit a first-order magnetostructural phase transition (FOMPT) generally exhibit large MCE.However, the hysteresis loss associated with the FOMPT can inhibit the MCE appropriateness of a material. 15,16Keeping these considerations in mind, there is an interest in minimizing the drawbacks of the FOMPT materials while maintaining the inherent MCE benefits.
Here we report an experimental study of the structural, magnetic, electronic transport, and magnetocaloric properties of the Ni 2 Mn 0.55 Co x Cr 0.45-x Ga (0 ≤ x ≤ 0.20) Heusler alloy system.We chose

EXPERIMENTAL TECHNIQUES
The Ni 2 Mn 0.55 Co x Cr 0.45-x Ga (0 ≤ x ≤ 0.25) samples were fabricated by the standard arc melting technique.The constituent elements (more than 3N purity) were used as starting materials.The ascasted ingots were annealed in vacuum at 1123 K for three days followed by quenching in cold water.The structure and phase purity of the compounds were assessed using a Scintag PAD X powder x-ray diffractometer (XRD), a device which employs Cu-Kα radiation and a theta-theta collection geometry.The XRD powder patterns were indexed and refined using the PowderCell analysis suite. 19he compositional purity of the compounds was explored using the backscattered electron detector of a Zeiss Supra 35 FEG scanning electron microscope (SEM) with an energy-dispersive x-ray spectroscopy (EDS) microprobe.The magnetization measurements were conducted on a Physical Property Measurement System (PPMS).∆S M as a function of temperature and field was determined by collecting M(H) data isothermally as a function of increasing field strength near each sample's respective transition temperature, and applying the expression: 20 The electrical resistivity as a function of temperature [ρ(T )] was collected in zero magnetic fields, in a lab-constructed four-probe cryostat unit in the temperature range of 77-500 K. Differential scanning calorimetry (DSC) curves were collected using a TA Instruments DSC2000 unit near each sample's respective FOMPT temperature in zero magnetic field.

RESULTS AND DISCUSSIONS
XRD patterns for the Ni 2 Mn 0.55 Co x Cr 0.45-x Ga (x = 0, 0.10, and 0.20) materials are shown in Figure 1.The austenitic L2 1 cubic and tetragonal martensite structures are equally present in the sample with x = 0.This coexistence is likely due to the near-room temperature FOMPT. 21For the cubic phase of the x = 0 sample a lattice parameter of a = 5.7629 Å was obtained, whereas the tetragonal phase exhibited lattice parameters of a = b = 5.3796 Å and c = 6.4248Å.The sample with x = 0.10 also exhibits elements of the same structural phases, but with a preference towards the martensite phase.For the cubic phase of the x = 0.10 sample, a lattice parameter of a = 5.7933 Å was obtained while lattice parameters of a = b = 5.3801 Å and c = 6.4700Å were retrieved for the tetragonal martensite phase.The sample with x = 0.20 exhibits a dominant martensite phase, with refined lattice parameters of a = b = 5.4015 Å and c = 5.5434 Å.The variation in lattice parameters as Cr is further replaced with Co may arise from the different metallic radii assumed by Cr and Co. 22 These variations in lattice parameters may be culpable for any changes in T C or T M observed.From the SEM micrographs shown within the insets of Figure 1, a sporadic, non-uniform density of dark spots is observed for samples with x ≤ 0.10.EDS revealed that while the light regions suitably match the target stoichiometry, the dark spots are almost entirely (>94 %) Cr, as has been observed in other off-stoichiometric Cr-rich Heusler alloys. 17 M(T ) curves of selected Ni 2 Mn 0.55 Co x Cr 0.45-x Ga (x =0, 0.05 and 0.10) materials, collected in a 1 kOe applied field, are shown in Figure 2.For each sample, a zero-field-cooled warming (ZFC) and field cooling curve (FC) was collected. 24Samples with x < 0.2 exhibit a FOMPT, signified by a sharp drop in magnetization accompanied by a thermal hysteresis.For x ≤ 0.10, both T C values increase linearly with x, such that the samples with x = 0 and 0.10 exhibit T C ZFC /T C FC values of 248 K/244 K, and 293 K/278 K, respectively.For x = 0.15, T C ZFC and T C FC decrease to 283 K and 274 K, respectively.Finally, for the samples with x = 0.20 and 0.25, the coinciding T C values decrease from 284 K to 282 K.
It should be noted that for x < 0.10, data collected under both the ZFC and FC protocols during the phase transition demonstrate a discontinuous character, emblematic of first-order phase transitions, resulting in a symmetrically-balanced hysteretic region (shaded region in Figure 2 for x = 0).For samples with x = 0.10 and 0.15, the asymmetric hysteretic region (see Figure 2 for the x =0.10 sample) occurs because the ZFC M(T ) curves exhibit first-order characteristics, whereas data collected under the FC protocol exhibits a second-order character.This suggests that the structural and magnetic phase transitions have become partially-decoupled in the samples with x = 0.10 and 0.15, such that while T M T C , they are in close enough proximity that some FOMPT characteristics are maintained.The absence of thermal hysteresis in the M(T ) data for x > 0.2 (not shown here) indicates that T C and T M have become fully-decoupled.
Normalized ρ(T ) measurements for selected samples are shown in Figure 3.The T M values for the martensite-to-austenite transition (T M M→A ) as determined from the data for the samples with x = 0, 0.1, and 0.2, are 244 K, 317 K, and 340 K, respectively.Furthermore, it is noted that the magnitude of ∆ρ does not vary significantly across the system, even though the degree of compositional disorder does.For the sample with x < 0.1, T C and T M values are close enough to form a FOMPT.However, for x =0.10 the T C and T M values diverge but are still within ± 10% of each other, meaning that a partially-decoupled FOMPT is formed.
Having observed varyingly-coupled FOMPTs near room temperature, there was justification for an assessment of this system's MCE properties.∆S M (T ) curves for selected samples were calculated for several field changes, as shown in Figure 4(a-b).For samples with x ≤ 0.15, the peak ∆S M (T ) is attained at the previously-determined FOMPT temperatures; for x > 0.15, the peak value occurs at T C .As expected, samples with a coupled FOMPT (x ≤ 0.05) have larger ∆S M values than samples with a decoupled FOMPT (x ≥ 0.20).For a field change of 50 kOe, the samples with x = 0, 0.1, and 0.2 exhibit peak ∆S M values of ∼11, ∼7, and ∼2 J/kg.K, respectively.While this may suggest that Co-substitution inhibits MCE potential, peak ∆S M may not be the most appropriate metric by which to adjudicate MCE performance, given that it doesn't account for hysteresis losses.From M(H) data collected as a function of both increasing and decreasing field strength in Figure 4(c-d), it is apparent that partial-decoupling of the FOMPT is accompanied by the elimination of magnetic hysteresis.As the difference between T C and T M values is somewhat larger when x = 0.10 than when x = 0, it is plausible that even a 50 kOe magnetic field may be unable to induce the competing magnetostructural phases that give rise to magnetic hysteresis.This reduction could mean that samples with partiallycoupled FOMPTs may be more appropriate for MCE applications, despite the comparatively-smaller ∆S M values.
To assess the repeatability of the discontinuous volume change experienced during FOMPT, a twenty-cycle DSC curve was collected for the samples with x = 0 and x = 0.10 in zero applied magnetic field, as shown in Figure 5.In both cases, cyclic repeatability indicates that the FOMPT is excellently reproducible on this time scale. 25

CONCLUSION
In summary, we have performed an experimental study of the Ni 2 Mn 0.55 Co x Cr 0.45-x Ga (0 ≤ x ≤ 0.25) Heusler alloy system.It was found that the replacement of a small fraction of Cr with Co can result in the partial decoupling of the coupled first-order magnetostructural phase transition exhibited by the parent compound.The decoupling eliminates the magnetic hysteresis losses.Electronic transport measurements indicate that all samples exhibit similar changes in resistivity during the structural phase transition.Furthermore, cyclic calorimetric measurements indicate that the structural phase transition is excellently-reproducible for both coupled and partially-decoupled first-order magnetostructural phase transitions.

FIG. 1 .
FIG. 1. Room temperature XRD patterns of selected Ni 2 Mn 0.55 Co x Cr 0.45-x Ga (x =0, 0.10, and 0.20) Heusler alloys.M and A denote indices corresponding to the tetragonal martensite and L2 1 cubic structures, respectively.The insets show SEM images of the corresponding samples (mag.bar = 8 µm).

FIG. 4 .
FIG. 4. (a) ∆S M (T ) of the Ni 2 Mn 0.55 Co x Cr 0.45-x Ga the x = 0 and (b) x = 0.10 Heusler alloy samples, for field changes of 20 kOe and 50 kOe.(c) M(H) of the x = 0 and (d) x = 0.1 samples, collected at a variety of temperatures near each sample's respective FOMPT temperature.