This site uses cookies. By continuing to use this site you agree to our use of cookies. To find out more, see our Privacy and Cookies policy.
Brought to you by:
Paper

Magnetic and transport properties of Zr1−xNbxCo2Sn

, , , , and

Published 23 April 2019 © 2019 IOP Publishing Ltd
, , Citation Meng Yang et al 2019 J. Phys.: Condens. Matter 31 275702 DOI 10.1088/1361-648X/ab161e

0953-8984/31/27/275702

Abstract

ZrCo2Sn is a potential candidate as a Weyl semimetal with a ferromagnetic ground state, and Nb-doping is expected to shift the Weyl points to the vicinity of Fermi level. We successfully synthesized a series of Zr1−xNbxCo2Sn single crystals with various concentrations of Nb (x  =  0, 0.1, 0.2, 0.275, 0.4, 0.5). All samples have a spinel structure and the lattice constant decrease as the Nb doping level increases. The magnetization and transport measurements suggest that the ferromagnetic ordering temperature can be strongly modified by the Nb doping. When x increases, the Curie temperature decreases significantly, accompanied by a change from metal-like to semiconductor-like behavior. There is a crossover for positive to negative MR at a temperature between 30 K to 50 K. In constant, the magnitude of the anomalous Hall resistance increases monotonously with decreasing temperature.

Export citation and abstract BibTeX RIS

1. Introduction

The Dirac and Weyl semimetals (DSM and WSM) have attracted enormous research interest in recent years. For DSM and WSM, the conduction band overlaps with valance band at certain points in the momentum space [13]. The DSMs with four-fold degenerate energy nodal points, such as Na3Bi [4, 5] and Cd3As2 [6, 7], have been successfully predicted and experimentally confirmed. When the inversion symmetry (IS) or time-reversal symmetry (TRS) is broken, the four-fold degenerate Dirac fermions will split into a pair of two-fold degenerate Weyl nodes. This mechanism results in two kinds of WSMs: IS-breaking WSMs, such as the TaAs family [810], and WTe2 [11, 12], and TRS-breaking WSMs, such as Y2Ir2O7 [2], HgCr2Se4 [13], and some Heusler alloys [1416]. Although the IS-breaking WSMs have been well characterized in the TaAs family, the evidence of the TRS-breaking WSMs is still scarce [1720].

The Heusler alloys, a large family with a vast class of materials comprising more than 1000 compounds [16, 2127], which exhibit a rich variety of electronic and magnetic properties, are also expected to host same candidates for WSMs without TRS. For example, evidence for WSM phase in GdPtBi [15] and Co3Sn2S2 [17] has been reported. Recently, a new series of Co-based Heusler materials have been predicted to be WSMs with TRS breaking [14, 28]. Among them, ZrCo2Sn alloy is interesting because it only posses one pair of Weyl points. However, these Weyl points are located 0.6 eV above the Fermi level. Doping Nb in the Zr site has been proposed to shift the two Weyl nodes to the Fermi level. Based on a theoretical calculation, with a Nb-doping concentration of 0.275, the Weyl points can be moved to the Fermi level. Although there have been some experimental efforts reported on the Co-based Heusler alloys [29, 30], the Nb-doped single crystals have not been synthesized and investigated, to the best of our knowledge.

In this study, we successfully synthesized a series of single crystals of a magnetic Heusler alloy, Zr1−xNbxCo2Sn, via a well-designed two-step method. The structural and physical properties were systematically investigated, especially for x  =  0.275. We found that with increasing the Nb doping level, the lattice constant decreases, which is accompanied with increase in resistivity, decrease in ferromagnetic phase transition temperature, and decrease in saturated magnetization. A crossover of the magnetoresistance (MR) from positive to negative was observed near 30–50 K. Additionally, a large anomalous Hall effect (AHE) was observed in this series of single crystals.

2. Experimental detail

Single crystals of Zr1−xNbxCo2Sn were grown from excess tin flux in two steps. High-purity Zr, Nb and Co (slug, 99.99%) at a molar ratio of 1  −  x: x: 2 were melted in an arc furnace filled with pure argon for several times. The weight loss was less than 1%. Then, the Zr1−xNbxCo2 alloy and excess tin (pill, 99.999%) were placed into an alumina crucible at a ratio of 1: 10 and sealed in a tantalum tube. The tantalum tube was then sealed in a quartz tube under high vacuum. After that, the quartz tube was heated to 1423 K over 10 h and dwelled for 5 h. Next, the tube was slowly cooled to 1173 K at a rate of 2 K h−1, followed by separating samples from flux in the centrifuge.

To determine the crystalline structure of this series of doping compounds, the single crystals with various doping concentrations were grounded into powder to perform the x-ray diffraction (XRD) measurement via Shimadzu, XRD-7000 detector by using Cu 1 radiation (λ  =  1.5418 Å). The lattice parameters were deduced by analyzing the obtained XRD patterns for each doping concentration. In order to further confirm the crystalline structure, a single crystal for x  =  0.275 was measured by Bruker D8 Venture High-Resolution Four-Circle Diffraction using Mo Kα1 radiation (λ  =  0.710 73 Å). Via the SHELXL-2014/5 program, a full-matric least-squares of F2 was used to refine the single-crystal XRD data [31]. The chemical composition of the crystals was determined by energy-dispersive x-ray spectroscopy with a Hitachi S-4800 scanning electron microscope at an accelerating voltage of 15 kV. Before the magnetic and transport measurements, some technical modifications were performed on the single crystals. To ensure that the electric current was more homogeneous in the crystals during transport measurement, we polished the single crystals into sheets with a thickness of less than 0.2 mm.

The magnetic properties were measured for all doping single crystals in a Magnetic Properties Measurement System (MPMS, Quantum Design Inc.) with a SQUID-VSM option. The magnetic susceptibility (χ) was measured between 2 K and 300 K in an applied field of H  =  1 kOe with the field-cooling (FC) and zero-field-cooling (ZFC) modes. Isothermal magnetization (MH) was measured at several fixed temperatures with a sweeping applied field from  −50 kOe to 50 kOe. The longitudinal resistivity ρxx at various applied fields of H  =  0, 50, 90 kOe was obtained in the Physics Property Measurement System (PPMS, Quantum Design Inc.) in a configuration of the four-terminal method. Four platinum wires were fixed on the polished crystal via silver epoxy. The MR and Hall resistivity ρxy were measured in a sweeping field from  −50 kOe to 50 kOe at various temperatures in a Janis He-4 system by a six-terminal contact geometry. A standard frequency lock-in technique was used during the measurement. The MR was defined as ${\rm MR}(\%)=({{R}_{H}}-{{R}_{0}})/{{R}_{0}}\times 100\%$ , where R0 and RH are the resistances without and with an applied field, respectively. The data of the MR and Hall resistivity were managed by symmetric/antisymmetric processing.

3. Result and discussion

3.1. Crystal structure

The refinement of the single-crystal XRD study illustrates that Zr0.725Nb0.275Co2Sn crystallizes in a cubic structure with space group Fm-3m (#225) and the lattice parameter a  =  6.097 50 Å. The further details are summarized in table 1. A schematic drawing of the crystal structure is shown in figure 1(a), where the Zr/Nb and Sn atoms form a NaCl-type frame with a simple cube of Co atoms inserted. The position of the Zr atoms (shown as green spheres) are partly occupied by Nb atoms (shown as gray spheres). The proportion of Nb atoms is approximately 30%, slightly higher than expected.

Table 1. Atomic coordinates and equivalent isotropic thermal parameters of Zr0.725Nb0.275Co2Sna.

Site Wyckoff position x y z Occup. Ueq
Sn 4b 0.500 00 0.000 00 0.000 00 1.000 0.0192 (16)
Zr 4a 1.000 00 0.500 00 0.500 00 0.7 (4)  0.022 (2)
Nb 4a 1.000 00 0.500 00 0.500 00 0.3 (4)  0.022 (2)
Co 8c 0.750 00 0.250 00 0.250 00 1.000  0.030 (2)

aSpace group: Fm-3m(#225); lattice constants: a  =  b  =  c  =  6.097 50 Å, α  =  β  =  γ  =  90°; unit-cell volume  =  226.702 022 Å3; d  =  9.619 g cm−3; Z  =  1; R  =  3.04 (25)%, RW2  =  6.89 (25)%, S  =  1.365, Npar  =  6.

Figure 1.

Figure 1. (a) Crystal structure of Zr1−xNbxCo2Sn. Green, gray, blue, and pink balls represent the Zr, Nb, Co, and Sn atoms, respectively. (b) The XRD patterns of Zr1−xNbxCo2Sn. The unmatched peaks with red asterisks are caused by Sn. The inset is a photograph of Zr1−xNbxCo2Sn single crystals. (c) The lattice parameter a as a function of x.

Standard image High-resolution image

The powder XRD patterns of the grounded single crystals for the five samples are presented in figure 1(b). The sharp diffraction peaks indicate the high crystalline quality, and the several additional low-density peaks marked with red asterisks correspond to the impurity of the Sn flux. The consistent peaks imply that the crystal structure remains in the Heusler phase of space group Fm-3m with Nb doping. The diffraction peaks shift to high degree with increasing x, obviously at a high degree, suggesting a change in the lattice parameter. The lattice constant a, refined from the powder XRD data, as a function of the doping concentration of Nb x, is plotted in figure 1(c). The value of a decreases monotonously as x increases, which may be related to the smaller atomic radius of the Nb atom (1.46 Å) compared to the Zr atom (1.6 Å). It must be noted that the nonstoichiometry of the Co atoms in these compounds is less than 2 (~1.83) and the vacancies of Co atoms are on the same level in all Zr1−xNbxCo2Sn samples. As a result, the lattice parameter for x  =  0 is 6.200 78 Å smaller than that found in previous reports [30, 32].

3.2. Magnetic properties

Figure 2(a) shows temperature-dependent magnetic susceptibility (χ) from 2 K to 300 K with an applied field of H  =  1 kOe for all samples. Only the results of the ZFC mode are presented because the data overlap with those obtained with the FC mode. The χT curves exhibit similar characteristics, indicating a ferromagnetic to paramagnetic phase transition for each of samples. The sharp drop in χ near the transition allows for a reliable extraction of the Curie temperature [33]. TC shifts a lower temperature with increasing x, which is accompanied with the reduction of the lattice constant. The magnitude of magnetic susceptibility at 2 K also drops considerably as the Nb concentration increases, further suggesting the weakening ferromagnetism with increasing Nb-doping. According to a previous theoretical work [14], the ferromagnetism in ZrCo2Sn originates from a short-range ferromagnetic exchange interaction between Co atoms. Even though in many materials the reduction in the lattice constant enhances in the ferromagnetic exchange interaction and thus the Curie temperature. Our data, namely the stronger ferromagnetism at large lattice constant, rather suggest that mechanism of ferromagnetic exchange interaction is more complicated.

Figure 2.

Figure 2. Temperature dependence of (a) the ZFC magnetic susceptibility χ and (b) its inverse χ−1 with an applied field H  =  1 kOe of every proportion. At high temperature, the inverse susceptibility can be fitted by the formula ${{\chi}^{-1}}=T/C-{{T}_{\theta}}/C$ .

Standard image High-resolution image

In the previous reports, the TC of polycrystalline samples ZrCo2Sn is approximately 468 K [32], 352 K higher than that of NbCo2Sn polycrystals [28]. In our experiment, the ZrCo2Sn single-crystals exhibit a much lower TC (~220 K), and the TC of Zr1−xNbxCo2Sn single crystals is lower than that of NbCo2Sn polycrystals when x  ⩾  0.275. Kushwaha et al [29] have recently proposed that the Co vacancies may be account for the change in TC of ZrCo2Sn single crystals.

Figure 2(b) illustrates the inverse susceptibility (χ−1) as a function of temperature. The data above the phase-transition temperature was fitted by the Curie–Weiss law via the formula ${{\chi}^{-1}}=T/C-{{T}_{\theta}}/C$ , where Tθ and C are the Weiss temperature and Curie constant, respectively. A deviation between the χ−1  −  T curves and fitted lines (dashed line in figure 2(b)) are observed. This deviation is fairly similar to that of HgCr2Se4 [34, 35], in which the spin correlations play an important role.

Figure 3(a) shows the isothermal magnetization as a function of the applied fields at T  =  2 K for different amounts of Nb doping. All samples exhibit very narrow hysteresis loops with a coercivity field of about 50 Oe. The magnetization reaches saturation at a field lower than H ~ 1 kOe. The value of the saturated magnetization (Msat) decreases monotonously from 0.87 to 0.45 µB as x is increased from 0 to 0.5. It is noteworthy that Msat of the ZrCo2Sn sample is only 0.87 µB, much lower than 1.56 µB, reported for a polycrystalline sample [36]. This suggests that the magnetic properties of (Zr,Nb)Co2Sn are very sensitive to the microscopic details of the samples, which deserve further investigation.

Figure 3.

Figure 3. (a) The magnetization of (Zr,Nb)Co2Sn as a function of the applied fields for 2 K at different Nb-doping levels. (b) Magnetization plotted as a function of µ0H at varying temperature with x  =  0.275.

Standard image High-resolution image

Figure 3(b) displays the M–H curves for the sample with x  =  0.275 at several temperatures (2–200 K). At 2 K, the magnetization reaches the maximum value of 0.68 µB/F.U. Above 140 K, the magnetization is almost linear with the applied fields, consistent with a transition to the paramagnetic phase at T  =  110 K. The overall behavior is very similar to HgCr2Se4 [34, 35], which is also a soft ferromagnet.

3.3. Electronic transport properties

Figure 4(a) displays the temperature dependences of the zero-magnetic field longitudinal resistivities of the Zr1−xNbxCo2Sn samples. The overall trend is that the increase in the Nb-concentration results in larger resistivity value, but the value of ρ rises by less than an order magnitude as x is varied from 0 to 0.5. A common feature of all ρT curves is that a hump-like feature appears the Curie temperature, which is marked with a downward arrow. The undoped sample (x  =  0) exhibits a metallic behavior at a wide temperature range below TC (~220 K). The Nb-doping leads to the shrinking of temperature range for the metallic behavior, which can be attributed to the decreasing TC in as well as the resistivity upturn at low temperatures. The resistivity upturn becomes more pronounced with increasing Nb-concentration. For the sample with x  =  0.5, the metallic behavior below TC nearly vanishes. It is also noteworthy that for x  =  0.275, the resistivity upturn is terminated with a small, but sharp drop as the temperature is lowered to 3 K. This feature might be attributed to a small amount of Sn phases in the sample, since the superconductivity transition temperature is 3.7 K and the resistivity drop can be suppressed by applying magnetic field.

Figure 4.

Figure 4. (a) The temperature dependence of the longitudinal resistivity ρ of Zr1−xNbxCo2Sn without applied fields. (b)–(e) T-dependence of ρ with µ0H  =  0, 5, 9 T for x  =  0, 0.275, 0.4, 0.5.

Standard image High-resolution image

Similar humps around TC in the ρT curves have been observed in many colossal magnetoresistance (CMR) materials (Ti2Mn2O7, HgCr2Se4) and Heusler compounds (Fe3Si, Fe2Val, Ru2CrSi) [3744, 35]. In the CMR materials, the nanoscale phase separation or the formation of the magnetic polarons has been widely accepted as the origin of the resistivity hump around TC [3739]. The strong scattering caused by these spin in homogeneities can be suppressed by applying magnetic field, leading to the CMR effect observed in large variety of materials. Figures 4(b)(e) show that similar phenomena also takes place in Zr1−xNbxCo2Sn samples. The resistivity upturn at low temperatures may also be attributed to spin disorder due to the weakening ferromagnetism with increasing Nb concentration. Only for undoped samples, the strong ferromagnetism allows for a truly metallic ground state. This situation is quite different from (Ti,Sc)2Mn2O7 samples, in which Sc-doping only weakly influence the magnetic order but can vary the resistivity by several orders of magnitude [41]. The low-T resistivity upturn observed in the (Ti,Sc)2Mn2O7 samples with large Sc concentration (e.g. x  =  0.4) can be attributed to Anderson localization. In Zr1−xNbxCo2Sn, The Nb-doping only results in modest increase in the resistivity, so the resistivity upturn is unlikely to be caused by the localization of charge carriers.

The MR curves of four samples are depicted in figures 5(a)(c). At low temperatures, MR is mostly positive and shows no sign of saturation up to 5 T. Such a MR is a property of the metallic phase, which has been observed in many materials, both magnetic and non-magnetic. At high temperatures (above 30–50 K, which varies the Nb-doping level), the MR becomes negative. This can be attributed to the reduced spin-dependent scattering in applied magnetic fields. The crossover from positive to negative MR can be seen more clearly in figure 4(d), in which a crossover temperature of ~30 K can be determined from the detailed measurements between 15 and 35 K for the sample with x  =  0.275. At such a temperature, the sample is still in the ferromagnetic phase.

Figure 5.

Figure 5. (a)–(c) The MR versus field strength from  −50 kOe to 50 kOe at varying temperature for different amounts of doping Nb. In low fields, the dip at 2 K for the x  =0 and x  =  0.1 samples is caused by the Sn impurity. (d) The details of the Zr0.725Nb0.275Co2Sn single crystal of MR versus µ0H from 15 K to 35 K.

Standard image High-resolution image

Figure 6(a) shows the Hall resistivity (−ρxy) as a function of magnetic field µ0H for x  =0.275 at various temperatures. A large AHE signal is found for T below 200 K, with a sharp increase at low temperatures near zero field. With T increasing, the AHE weakens and the sharp increase no long exist. In general, the Hall resistivity of a magnetic compound mainly comes from two parts: normal Hall resistivity ($\rho _{xy}^{0}$ ) and anomalous Hall resistivity ($\rho _{xy}^{{\rm AHE}}$ ) due to the magnetism in the sample,${{\rho}_{xy}}=\rho _{xy}^{0}+\rho _{xy}^{{\rm AHE}}={{R}_{{\rm H}}}\times B+{{\mu}_{0}}{{R}_{{\rm S}}}\times M$ , in which ${{R}_{{\rm H}}}$ and ${{R}_{{\rm S}}}$ are respectively the normal and anomalous Hall coefficients, ${{\mu}_{0}}$ is the vacuum permeability, and M is the magnetization of the sample [45, 46]. $\rho _{xy}^{{\rm AHE}}$ increases rapidly and is dominant in ρxy at low fields. When the magnetization is saturated, $\rho _{xy}^{0}$ is responsible for the linear behavior of ρxy at high fields. Based on this, an electron density n  =  ~1.26  ×  1022 cm−3 can be extracted for the sample (x  =  0.275) at 2 K. Figure 6(b) further shows the normalized  −ρxy plotted as a function of normalized M at 50 K to 150 K. The data at all temperatures collapse nicely onto a single curve. This suggests that the Hall signal is dominated by the AHE and a unified description of magnetic and electronic structures can be developed for this wide temperature range.

Figure 6.

Figure 6. (a) The Hall resistivity  −ρxy as a function of magnetic field µ0H at varying temperature for 0.275 with I  =  2 mA. (b) The ρxy versus M from 50 K to 150 K for x  =  0.275.

Standard image High-resolution image

4. Conclusions

We have studied a series of Zr1−xNbxCo2Sn single crystals (x  =  0–0.5) grown via the Sn-flux method. No structural transition has been observed for the entire range of Nb concentration. The increase in the Nb-doping level causes a reduction in the lattice constant, which is accompanied by a significant drop in the Curie temperature and saturation magnetization. As a result, the electron transport properties are also modified greatly by the Nb-doping and can be mostly explained by the weakening ferromagnetism and metallic phase in this material system. Our work suggests that Nb-doping is probably not a good approach for achieving the Weyl semimetal phase in ZrCo2Sn because of the unwanted negative effect on the ferromagnetic order.

Acknowledgment

We thank Tian Qian and Hongming Weng for valuable discussion. This work was supported by the National Key Research and Development Program of China (Grants No. 2017YFA0302901 and No. 2016YFA0300604), the National Natural Science Foundation of China (Grants No. 11774399 and No. 61425015), Beijing Natural Science Foundation (Grant No. Z180008), and the Chinese Academy of Sciences (Grant QYZDB-SSW-SLH043).

Please wait… references are loading.
10.1088/1361-648X/ab161e