2024 Volume 33 Issue 7
Article Contents

Jin-Jing Liang, Xue-Kui Xi, Wen-Hong Wang, Yong-Chang Lau. Tailoring-compensated ferrimagnetic state and anomalous Hall effect in quaternary Mn–Ru–V–Ga Heusler compounds[J]. Chinese Physics B, 2024, 33(7): 077504. doi: 10.1088/1674-1056/ad34c8
Citation: Jin-Jing Liang, Xue-Kui Xi, Wen-Hong Wang, Yong-Chang Lau. Tailoring-compensated ferrimagnetic state and anomalous Hall effect in quaternary Mn–Ru–V–Ga Heusler compounds[J]. Chinese Physics B, 2024, 33(7): 077504. doi: 10.1088/1674-1056/ad34c8

Tailoring-compensated ferrimagnetic state and anomalous Hall effect in quaternary Mn–Ru–V–Ga Heusler compounds

  • Received Date: 22/01/2024
    Accepted Date: 07/03/2024
    Available Online: 01/06/2024
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(8)

Article Metrics

Article views(293) PDF downloads(0) Cited by(0)

Access History

Tailoring-compensated ferrimagnetic state and anomalous Hall effect in quaternary Mn–Ru–V–Ga Heusler compounds

Abstract: Cubic Mn2RuxGa Heusler compound is a typical example of compensated ferrimagnet with attractive potential for high-density, ultrafast, and low-power spintronic applications. In the form of epitaxial thin films, Mn2RuxGa exhibits high spin polarization and high tunability of compensation temperature by freely changing the Ru content x in a broad range (0.3 < x < 1.0). Herein Mn–Ru–Ga-based polycrystalline bulk buttons prepared by arc melting are systematically studied and it is found that in equilibrium bulk form, the cubic structure is unstable when x < 0.75. To overcome this limitation, Mn–Ru–Ga is alloyed with a fourth element V. By adjusting the content of V in the Mn2Ru0.75VyGa and Mn2.25–yRu0.75VyGa quaternary systems, the magnetic compensation temperature is tuned. Compensation is achieved near 300 K which is confirmed by both the magnetic measurement and anomalous Hall effect measurement. The analyses of the anomalous Hall effect scaling in quaternary Mn–Ru–V–Ga alloy reveal the dominant role of skew scattering, notably that contributed caused by the thermally excited phonons, in contrast to the dominant intrinsic mechanism found in many other 3d ferromagnets and Heusler compounds. It is further shown that the Ga antisites and V content can simultaneously control the residual resistivity ratio (RRR) as well as the relative contribution of phonon and defect to the anomalous Hall effect a″/a′ in Mn–Ru–V–Ga, resulting in a scaling relation a″/a′ ∝ RRR1.8.

1.   Introduction
  • Ferrimagnets, which consist of inequivalent antiferromagnetically coupled magnetic sublattices, have become one of the most attractive material classes in spintronics[1–7] because they combine the advantages of ferromagnets and antiferromagnets. On one hand, the ferromagnet-like electrical properties of ferrimagnets can be more easily used to achieve electrical reading and writing, compared to antiferromagnets.[3] On the other hand, ferrimagnets show antiferromagnet-like dynamics and they may switch much faster than ferromagnets.[7] Moreover, inequivalent magnetic sublattices exhibit different temperature dependence, so that in some cases, a full magnetic compensation may appear in a so-called compensated ferrimagnet. The temperature at which the net magnetization vanishes is the magnetic compensation temperature (TM). Compensated ferrimagnets which show no stray fields, high spin polarization, large anomalous Hall effect, and low damping at room temperature have attractive potential applications in high-density, ultrafast, and low-power spintronics.[8–16]

    The common ferrimagnets include rare-earth iron garnets,[17–19] rare-earth transition metal alloys,[20–22] Mn4N-based nitrides,[23,24] and some Mn-based Heusler compounds.[25–31] By varying the composition, not only can the compensated ferrimagnetic state be achieved, but also TM can be regulated. Cubic Mn2RuxGa thin film is a typical example of compensated ferrimagnet with high spin polarization and high tunability of TM by freely varying the Ru content x in a broad range (0.3 < x < 1.0).[27] So far, Mn–Ru–Ga alloy has mostly been realized in the form of thin films, such as epitaxial growth on MgO substrates,[27,30,32] which is known to be a popular method of stabilizing metastable phases of matter. It remains unclear to what extent is the cubic Mn–Ru–Ga ternary alloy thermodynamically stable and how the atomic order evolves as a function of x. Another important feature of the compensated ferrimagnet is the emergence of finite anomalous Hall effect, even when the net magnetization exactly vanishes. The existing compensated ferrimagnetic thin films, however, typically show mediocre change of the longitudinal resistance with temperature[27,33] and complex anomalous Hall effect loops[34–37] which impede accurate extraction of the anomalous Hall effect scaling. The aforementioned issues motivate us to study the Mn–Ru–Ga-based ternary alloys and Mn–Ru–V–Ga quaternary alloys in bulk form.

    We find that the cubic Heusler structure of Ru-deficient Mn2RuxGa and Mn3−xRuxGa bulk alloys becomes unstable for x < 0.75. Alloying with vanadium stabilizes the cubic structure of Mn–Ru–V–Ga and enables tunable compensated ferrimagnetic state in a wider temperature range, including room temperature. Moreover, although they are a ternary or quaternary alloy system, our polycrystalline bulk samples exhibit appreciable residual resistivity ratio (RRR) ranging from 2 to 5 (compared to previous thin film samples), allowing us to comfortably study the scaling relation of the anomalous Hall effect. We find nonnegligible thermally activated phonon contribution to the extrinsic anomalous Hall effect in these compensated ferrimagnets and establish a link between the RRR and the relative contribution of phonon and defect to the anomalous Hall effect.

2.   Experimental techniques
  • Polycrystalline buttons of Mn2RuxGa, Mn3−xRuxGa, Mn2Ru0.75VyGa, and Mn2.25–yRu0.75VyGa were prepared by arc melting. Extra 5% of Mn has been added to compensate for the losses. Samples were repeatedly melted four times after flipping to ensure the homogeneity. Mn2Ru0.75VyGa and Mn2.25–yRu0.75VyGa (0 ≤ y ≤ 0.25) samples were sealed inside quartz tubes evacuated to 2 × 10−4 Pa, annealed at 1073 K for 7 days, and then cooled to room temperature for one day to improve the atomic ordering. The crystal structure was determined by the powder x-ray diffraction (XRD). The density was measured by the Archimedes drainage method. Magnetic properties were measured by using the Quantum Design magnetic property measurement system in SQUID-VSM mode. Magnetotransport properties were measured by using standard 6 contact method on thin slab samples with a typical thickness of ∼ 100 μm in a physical property measurement system.

3.   Results and discussion
  • Figure 1(a) illustrates the crystal structure of the stoichiometric and atomically-ordered Mn2RuGa. Ru occupies 4d sites, Ga is on 4b sites and Mn occupies inequivalent 4c and 4a sites, thus forming the noncentrosymmetric Hg2CuTi-type inverse Heusler structure (Space group No. 216).[38] For quaternary Mn–Ru–V–Ga alloy, small amount of V with stronger affinity to Ga is expected to preferably replace some of the Mn atoms on 4a sites. The magnetism of the compound is mainly governed by the antiferromagnetically coupled 4a and 4c Mn sublattices that are inequivalent in terms of chemical environment and atomic occupancy.[29] The basic concept of the magnetization compensation in a compensated ferrimagnet that consists of antiferromagnetically coupled M1 and M2 magnetic sublattices is depicted in Fig. 1(b). The inequivalent M1 and M2 sublattices exhibit different total moment M at 0 K and dissimilar temperature dependency. As the temperature increases, if the sublattice with smaller M at 0 K (here M1) shows weaker temperature dependence than M2 sublattice, magnetic compensation may occur at the crossover TM where the magnetizations of M1 and M2 sublattices are equal in magnitude but opposite in sign. Below TM, the net moment of the compound is governed by the M2 sublattice, whereas above TM, M1 sublattice is dominant. In order to adapt a given application, it is desirable to be able to regulate the TM over large temperature range. In the following subsections, such an ability will be demonstrated for quaternary Mn–Ru–V–Ga compound.

  • We first study ternary Mn–Ru–Ga-based polycrystalline buttons Mn2RuxGa and Mn3−xRuxGa in the form of equilibrium bulk. The nominal compositions of the former series are similar to that studied by Kurt et al.[27] in the form of epitaxial thin films. This series is considered off-stoichiometric because the Ga concentration is higher than the ideal 25 at.% for a full Heusler compound of general formula X2YZ (Z being a main group element). For comparison, we also prepare the latter sample series with nominal Ga concentration fixed at 25 at.%. In Figs. 2(a) and 2(b), we show the powder XRD of as-cast Mn2RuxGa and Mn3−xRuxGa. When x is less than 0.75, these alloys are no longer in a pure cubic phase. Peaks that correspond to the tetragonal and hexagonal phases appear in Mn2Ru0.5Ga, and the parasitic tetragonal phase emerges in Mn2.5Ru0.5Ga. Furthermore, we did not succeed in suppressing these parasitic phases by post annealing. We therefore conclude that the cubic structure is unstable for the two series when x < 0.75, which substantially reduces the tunability of bulk Mn–Ru–Ga ferrimagnet, notably TM cannot be tuned to room temperature.

    Referring to the Slater–Pauling rule,[39–41] the magnetic moment M at 0 K of an X2YZ half-metal with L21 full Heusler structure follows M = Nv – 24 (Nv is the number of valence electrons per formula). Alloying a fourth element (e.g., Ti, V, Cr) with less valence electrons than Mn and Ru into Mn2RuxGa (M = 1 μB/f.u. for x = 1) may extend the tuning range of TM. We choose and alloy Mn–Ru–Ga with V because Mn2VGa was known to be a cubic Heusler with a high Curie temperature of 783 K.[43] In contrast, since Mn2CrGa does not crystallize into cubic structure,[42] and there is, to the best of our knowledge, no experimental report on Mn2TiGa, alloying with Cr and Ti may not be as favorable as V in terms of phase stability. By fixing the Ru concentration at x = 0.75, we systematically investigate Mn2Ru0.75VyGa and Mn2.25–yRu0.75VyGa quaternary systems. Again, the former series is off-stoichiometric in terms of Ga, whereas the latter series nominally fixes 25 at.% of Ga.

    The powder XRD spectra for the two series after annealing are plotted in Figs. 2(d) and 2(f). All the samples crystallize in the cubic phase without showing noticeable parasitic phase. We observe strong reflections that can be indexed by the (hkl) Bragg peaks satisfying the relation h + k + l = 4n (n = integer). As shown in Fig. 2(g), the steady and monotonically decreasing trend of the extracted lattice parameter a with the increase of V composition y provides another evidence of systematic alloying. In addition, as shown in the enlarged XRD patterns for 2θ ranging from 23° to 32° [Figs. 2(c) and 2(e)], weak superlattice peaks corresponding to the (111) and (200) reflections are clearly visible. This mainly indicates that the position of the heaviest Ru atoms is at least partially ordered. Note that we do not intend to estimate the ordering parameter of these samples because additional assumptions must be made, that is, how other atoms are ordered and whether an atomic site is always fully occupied, with a total atomic occupancy always being equal to 1.

    In fact, the actual atomic ordering and occupancy of Mn2RuxGa are still controversial. In the pioneering work by Kurt et al.,[27] Mn2RuxGa was assumed to be an intermediate compound between the half Heusler and inverse Heusler structure, where Ru gradually fills the empty 4d site until x = 1. More recently, according to the thin film density estimated from the critical angle of the x-ray reflectivity data, Siewierska et al.[28] showed that all the investigated Mn–Ru–Ga films adopt full atomic occupancy with no obvious sign of vacancy. Herein, these Heusler compounds are successfully synthesized in the form of bulk buttons, enabling us to independently and directly measure the density of these samples by using the Archemede drainage method. For the three selected samples, their measured density is compared with the expected density in the cases of full atomic occupancy (full) and with atomic vacancy (vacant), as shown in Fig. 2(h). Experimental data support the scenario with full atomic occupancy in all cases, with and without V alloying. Next, we shall focus on the magnetic and magnetotransport characterizations of Mn2Ru0.75VyGa and Mn2.25–yRu0.75VyGa quaternary compounds for 0 ≤ y ≤ 0.25.

  • The magnetic properties of the off-stoichiometric Mn2Ru0.75VyGa for 0 ≤ y ≤ 0.25 are summarized in Fig. 3. The magnetization M measured on cooling with an applied magnetic field of 0.1 T is plotted against temperature T in Fig. 3(a). For all the samples, M values remain finite at 400 K, indicating that the Curie temperatures of these compounds is well above 400 K. As the temperature decreases, M first decreases to about zero and then increases. This is an indication of the magnetic compensation as explained in Fig. 1(b). To further confirm the compensation characteristics, we first apply a magnetic field of 0.5 T above the compensation temperature to magnetize the sample, and then remove the magnetic field to measure the temperature dependence of magnetization M(T) at zero magnetic field (i.e., in remanence) on cooling. These zero-field-cooled remanent M(T) curves are shown in the inset of Fig. 3(b), which directly confirm the magnetic compensation feature with the abscissa of the zero-moment point being the TM. The TM of all the measured samples is summarized in Fig. 3(b). Adding V consistently reduces TM from 380 K to 260 K. For y = 0.1 and y = 0.175, the TM is close to room temperature.

    The magnetic field dependence of magnetization M(H) for Mn2Ru0.75V0.1Ga (i.e., y = 0.1) measured at different temperatures are shown in Fig. 3(c). The inset shows the enlarged M–H loops in the low field region. The M–H loop at 300 K only shows a modest increase of the coercivity (< 0.1 T) due to the vanishing net magnetization. This is in contrast to the previous reports where biaxially-strained thin film samples grown on MgO (001) substrates tend to show huge or even diverging coercivity (> 9 T) near TM.[28] We infer that our annealed bulk samples are mostly relaxed with negligible tetragonal distortion and the weak magnetocrystalline anisotropy is solely provided by the cubic structure. Figure 3(d) summarizes the variations of magnetization at saturation Ms with y at 5 K and 300 K. As y increases, Ms at 5 K decreases monotonically, whereas Ms at 300 K first decreases and then increases due to the crossing of TM.

    The magnetic properties of the stoichiometric Mn2.25–yRu0.75VyGa series are shown in Fig. 4. Similar to Fig. 3, TM for this sample series can be obtained from the M(T) curves measured with an applied magnetic field of 0.1 T as shown in Fig. 4(a) or from the remanent M(T) curves measured under zero-field-cooled condition as plotted in the inset of Fig. 4(b). The extracted TM against V composition y is shown in Fig. 4(b). Again, TM tends to decrease with the increase of V content y. For y = 0 (i.e., Mn2.25Ru0.75Ga), the higher Mn content than Mn2Ru0.75Ga further enhances TM to above 400 K. To counterbalance this trend, a higher amount of V content (y = 0.2) is needed to achieve compensation near 300 K for this series.

    Figure 4(c) shows the magnetic field dependence of magnetization M(H) of Mn2.05Ru0.75V0.2Ga at different temperatures. The inset shows the enlarged M–H loops at low fields, again reflecting a weak enhancement of coercivity near TM. Figure 4(d) summarizes the variation of Ms with y at 5 K and 300 K. Similar to the TM, as y increases, Ms at 5 K and 300 K first decreases, then it slightly increases (for y ∼ 0.15), and finally it decreases again. The nonmonotonic change in the magnetic property with the increase of y (but not shown in the lattice constant [see Fig. 1(g)]) may be attributed to the preferred site occupancy of V that varies as y evolves, which in turn adjusts the site-specific occupancy of the two inequivalent, antiferromagnetically couple Mn sites. The magnetic compensation conditions are known to be very sensitive to such a change. Additional element-specific measurements, such as the anomalous XRD and the x-ray magnetic circular dichroism or the site-specific characterization such as the nuclear magnetic resonance are however required to verify this hypothesis.

    Figure 5(a) summarizes the evolution of Ms at 5 K with the total number of valence electrons, Nv for the two sample series. The solid symbol and open symbol represent data points obtained from the stoichiometric Mn2.25–yRu0.75VyGa and off-stoichiometric Mn2Ru0.75VyGa, respectively. Note that Nv is calculated based on the structural model with full atomic occupancy without vacancy as deduced from the density measurement. We find that Nv for all samples is greater than 24 and the data from the two sample sets can be combined and described by a linear fit. By extrapolating the linear fitting line, Ms is close to zero when Nv approaches 24, which is consistent with Slater–Pauling curve. However, the slope of the linear fitting line is 0.76±0.12 μB/f.u./electron (less than 1), suggesting that these quaternary alloys are not half-metals. Figure 5(b) summarizes the variation of TM with Nv. Overall, as Nv decreases, TM decreases from above 400 K to 260 K for Mn2Ru0.75V0.25Ga. The compensation at room temperature is realized by adjusting the V content and changing Nv to about 24.4. Furthermore, the data from the two sample series with different Ga atomic ratios are naturally reconciled when Nv deduced from the fully occupied structural model is considered.

  • The Hall loops of the Mn2Ru0.75V0.1Ga and the Mn2.05Ru0.75V0.2Ga thin slabs measured at different temperatures are shown in Figs. 6(a) and 6(b). The ρyx is the Hall resistivity measured with an applied magnetic field perpendicular to the slab plane. The sign reversal of the anomalous Hall effect provides another evidence that the magnetic compensation is indeed achieved near 300 K. The normal Hall coefficient strongly depends on the temperature, indicating that the Fermi level lies close to the band edge. Similar results are reported for Mn–Ru–Ga thin films.[27] The magnetoresistance (MR), defined as MR(μ0H) = (ρxx(μ0H) – ρxx(μ0H = 0)) / ρxx(μ0H = 0) where ρxx is the longitudinal resistivity, for Mn2Ru0.75V0.1Ga and Mn2.05Ru0.75V0.2Ga at different temperatures is shown in Figs. 6(c) and 6(d). At 5 K, the MR of Mn2Ru0.75V0.1Ga is negative within a magnetic field of 5 T while the MR of Mn2.05Ru0.75V0.2Ga is only negative at low fields and it becomes positive at higher fields. Near the compensation temperature, the signs of MR for both Mn2Ru0.75V0.1Ga and Mn2.05Ru0.75V0.2Ga reverses. Figures 6(e) and 6(f) show the absolute value of the anomalous Hall conductivity |σAH|=|ρyx|/(ρxx2+ρyx2) and anomalous Hall angle |θAH| = |ρyx|/ρxx, against temperature for the two sample. On the whole, |σAH| and |θAH| are rather smooth across TM, which is expected for an ideal compensated ferrimagnet where the sign reversal of σAH across TM is solely due to the change of the dominant magnetic sublattice below and above TM, without involving structural or phase transformation. Although we still observe a slight decrease of |σAH| and |σAH| near TM, most probably caused by minor spatial inhomogeneity of the alloy composition and TM, the inhomogeneity induced complex anomalous Hall effect previously found in epitaxial Mn–Ru–Ga[32] and amorphous Co–Gd[44] compensated ferrimagnetic thin films can be largely mitigated in our annealed bulk samples. Overall, we have shown that the compensated ferrimagnets can show finite anomalous Hall effect even when the net magnetization exactly vanishes at TM and ideally there should be no anomaly in |σAH| and |σAH| across TM.

  • To separate various contributions to the anomalous Hall effect in a magnetic conductor, a common practice is to study the scaling of σAH or anomalous Hall resistivity σAH with the longitudinal conductivity σxx=ρxx/(ρxx2+ρyx2) or longitudinal resistivity ρxx, which is considered to reflect the transport lifetime, τ.[45] For intrinsic anomalous Hall effect that arises from the Berry curvature of the band structures or the extrinsic anomalous Hall effect due to side jump mechanism, σAH is independent of σxx and τ (similarly ρAH is proportional to ρxx2). For extrinsic anomalous Hall effect due to skew scattering, σAH is proportional to σxx and τ (or ρAH is proportional to ρxx).

    The analysis can be further refined by discerning skew scattering contribution due to defects (dominant at low temperature) and thermally-excited quasiparticles, e.g., phonons and magnons (negligible at low temperature).[46] Let ρxx(T) = ρxx0 + ρxxT (T), then the general expression of ρAH will be

    where a′, a″, b′, and b″ are fitting parameters, ρxx0 = ρxx(T = 4 K) and ρxxT(T) = ρxx(T) – ρxx(T = 4 K) are the residual resistivity and the resistivity contribution that depends on temperature, respectively.

    Obviously, the accuracy of such a separation partly relies on our ability to experimentally tune the range of σxx and τ, by varying the temperature. Concerning this point, compounds and alloys in bulk form are advantageous compared with those in thin films because they typically possess better atomic ordering and less defect, leading to higher residual resistivity ratio, RRR = ρxx (T = 300 K)/ρxx(T = 4 K). In addition, the negligible interfacial scattering in bulk samples allows us to focus on the electronic transport that is inherent to the bulk properties of the material, and is not related to the interfacial effects that is inevitably involved and dependent on the adjacent layer. This is the reason why our Mn–Ru–V–Ga buttons possess significantly higher RRR than previous Mn–Ru–Ga thin films.[27]

    The variation of |σAH| with σxx for several Mn–Ru–V–Ga alloy samples of varying V content y is shown in Fig. 7(a). Solid and empty symbols represent the data for stoichiometric Mn2.25–yRu0.75VyGa and off-stoichiometric Mn2Ru0.75VyGa series respectively. On the whole, |σAH| is proportional to σxx with a nearly constant |σAH|∼ 1%, implying that the main contribution of anomalous Hall effect in Mn–Ru–V–Ga is extrinsic skew scattering mechanism. The dependence of anomalous Hall resistivity |ρAH| with ρxx2 are shown in Fig. 7(b). Obviously, |ρAH| is not linear with ρxx2, which again confirms the negligible contribution of the intrinsic mechanism, i.e., b″ is small. Figure 7(c) show the variation of |ρAH| with ρxxT. The |ρAH| shows excellent linear relationship with ρxxT, indicating that the contribution of thermally-excited phonons, a″, is significant. Note that since the temperature-dependent b″ is negligibly small in the Mn–Ru–V–Ga system, we have also dropped the b′ term, which sometimes is attributed to the extrinsic side jump mechanism. Therefore, a relatively simple phenomenological relation |ρAH| = a′ρxx0 + a″ρxxT is sufficient to describe all our experimental data. The fact that a′ and a″ are of the same sign suggests that the two skew scattering contributions are additive. For comparison, the anomalous Hall effect in epitaxial Fe,[46] Co[47] or L10–Mn1.5Ga films[48] was found to be governed by the intrinsic mechanism, and the phonon-induced skew scattering can be ignored. Non-negligible phonon skew scattering contribution was also reported in some Heusler alloys,[49–51] although we should note that the conductivity of these systems is in the “moderately dirty regime” with RRR that is close to 1 or even slightly less than 1. Therefore, we believe that it is unclear whether the changes in the anomalous Hall effect with temperature in these systems can be entirely attributed to the phonon-induced skew scattering.

    In order to further elucidate how the extrinsic skew scattering evolves with V content y, we plot and compare the RRR of Mn2Ru0.75VyGa and Mn2.25–yRu0.75VyGa series in Fig. 8(a). Since RRR closely relates to the defects and reflects the overall quality of the sample, it is therefore reasonable to assume that the RRR is correlated with the extrinsic skew scattering to a certain degree. The RRR of Mn2Ru0.75VyGa and Mn2.25–yRu0.75VyGa are shown in Fig. 8(a). Our first observation is that the stoichiometry of Ga (i.e., maintaining 25 at.% of Ga) has decisive implication on the RRR. For the two ternary alloys Mn2.25Ru0.75Ga (25 at.% of Ga) and Mn2Ru0.75Ga (26.7 at.% of Ga), RRR of the former is enhanced by more than a factor of 2 than that of the latter. The density measurement suggests that the excess Ga atoms occupy Mn or Ru sites, so we can conclude that these main group element antisites are mainly responsible for the reduction of RRR. A small amount of V first acts as additional scattering center, giving rise to the lowest RRR for y = 0.1 of the two series. Then, the RRR increases slightly to ∼ 3 upon increasing the V content to y = 0.25.

    Figures 8(b)–8(d) show the variation of a′, a″, and a″/a′ with RRR, respectively. In general, as RRR increases, a′ tends to decrease whereas a″ shows the opposite increasing trend. The relative contribution of the two skew scattering mechanisms against the RRR value can be phenomenologically described by an intriguing scaling relation a″/a′∝ RRR1.8 spanning over nearly one order of magnitude, as shown in Fig. 8(d). Interestingly, for off-stoichiometric Mn2Ru0.75VyGa series (red symbols), the a″/a′ ratio is typically less than 1, reflecting the dominant role of extrinsic skew scattering induced by Ga antisite defects, with appreciable y intercept in the |ρAH| versus ρxxT plot in Fig. 7(d). In contrast, when 25-at.% Ga stoichiometry is maintained, e.g., for Mn2.25–yRu0.75VyGa series (black symbols), the larger-than-unity a″/a′ ratio suggests that the thermally-excited phonons make significant contribution to the observed anomalous Hall effect. The observed scaling relation a″/a′∝ RRR1.8 implies that although a′ and a″ skew scattering contributions to the anomalous Hall effect in Mn–Ru–V–Ga quaternary alloys are sensitive to RRR and controllable by the composition as well as the atomic ordering, there may be a tradeoff between the two parameters, resulting in a rather constant anomalous Hall angle of ∼ 1%. To further enhance the anomalous Hall effect in compensated ferrimagnet, it would be highly desirable to utilize the collective contribution of the intrinsic and extrinsic anomalous Hall effects. Moreover, it would be interesting to test the validity and the universality of such a relation in other magnetic systems. One potential candidate is Fe3Sn2,[52] which is known to show large RRR and large extrinsic anomalous Hall effect at low temperature. However, there is an additional complication that the intrinsic contribution, the skew scattering contribution due to defects and that due to phonon should be carefully separated.

4.   Conclusions
  • In summary, we have systematically studied the compensated ferrimagnetic state and anomalous Hall effect scaling in ternary Mn–Ru–Ga and quaternary Mn–Ru–V–Ga-based polycrystalline Heusler compounds prepared by arc melting. Alloying with V stabilizes the cubic Heusler structure and substantially widens the tuning range of the magnetic compensation in these compounds. Detailed analyses of the anomalous Hall effect scaling reveal the dominant role of skew scattering mechanism, for which the contributions of Ga antisite defects and thermally-excited phonons are additive and controllable by varying the composition. These results provide a guideline to design compensated ferrimagnetic state with large anomalous Hall effect in Heusler compounds.

Figure (8)  Reference (52)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return