betaFeedback

WorksPhysics and Astronomy

Two types of colossal magnetoresistance with distinct mechanisms in Eu5In2As6

Abstract

Recent reports of colossal negative magnetoresistance (CMR) in a few magnetic semimetals and semiconductors have attracted attention, because these materials are devoid of the conventional mechanisms of CMR such as mixed valence, double exchange interaction, and Jahn-Teller distortion. New mechanisms have thus been proposed, including topological band structure, ferromagnetic clusters, orbital currents, and charge ordering. The CMR in these compounds has been reported in two forms: either a resistivity peak or a resistivity upturn suppressed by a magnetic field. Here we reveal both types of CMR in a single antiferromagnetic semiconductor Eu5In2As6. Using the transport and thermodynamic measurements, we demonstrate that the peak-type CMR is likely due to the percolation of magnetic polarons with increasing magnetic field, while the upturn-type CMR is proposed to result from the melting of a charge order under the magnetic field. We argue that similar mechanisms operate in other compounds, offering a unifying framework to understand CMR in seemingly different materials.

Introduction

Antiferromagnetic (AFM) materials with coupled charge, spin, and lattice degrees of freedom are coming to the forefront of electronic, spintronic, and caloritronic applications [1, 2, 3]. The Eu-based Zintl compounds are emerging as a fascinating family of functional AFM semiconductors with intertwined degrees of freedom. They exhibit exotic properties such as fluctuating Weyl nodes in EuCd2_2As2_2 [4, 5], axion insulator phase in EuIn2_2As2_2 [6, 7], and colossal magnetoresistance (CMR) in EuCd2_2P2_2 [8, 9, 10] and Eu5_5In2_2Sb6_6 [11, 12, 13, 14, 15]. Since CMR is particularly applicable in sensing and logic devices, it has been sought in a variety of materials including the chalcogenide Zintl phase Mn3_3Si2_2Te6_6 [16] and Dirac semimetals such as CeSbTe and EuAuSb [17, 18]. The general features of CMR in these materials are similar; they all exhibit a peak in their zero-field resistivity curves ρ(T)\rho(T) which is suppressed by an external magnetic field. In some cases, such as Eu5_5In2_2Sb6_6 and Mn3_3Si2_2Te6_6, a sharp upturn is observed in ρ(T)\rho(T) at low temperatures which is also suppressed by the magnetic field. Different mechanisms have been proposed for either type of CMR (peak or upturn), including ferromagnetic (FM) clusters [9], band structure reconstruction [19], chiral orbital currents [16], and interplay between Dirac fermions and charge ordering [17].

Here, we reveal both types of CMR in a single AFM semiconductor Eu5_5In2_2As6_6 and discuss the underlying mechanism for each type. A few properties of the title compound make it ideal for understanding CMR in the context of other related materials mentioned above. First, Eu5_5In2_2As6_6 is topologically trivial as shown by recent first-principles calculations [20], so Dirac fermions are irrelevant to its CMR, unlike the proposals for CeSbTe and EuAuSb [17, 18]. Second, Eu5_5In2_2As6_6 is a semiconductor with a band gap of 35 meV corresponding to temperature and field scales of about 400 K and 300 T. Thus, its sharp resistivity upturn at 15.5 K which is completely suppressed by 9 T, as shown here, cannot be due to a field- or temperature-induced transition from an insulator to a metal. Third, Eu2+^{2+} ions in Eu5_5In2_2As6_6 are not at the center of high-symmetry anionic polyhedra, so they are not subjected to orbital currents as proposed for Mn3_3Si2_2Te6_6 [16].

In this work, we argue that the peak-like CMR in Eu5_5In2_2As6_6 at high temperatures is related to polaron physics while the upturn-like CMR at low temperatures is potentially related to charge segregation or charge ordering. The extremely large magnitudes of both effects indicate a strong coupling among charge, spin, and lattice degrees of freedom, a property that could be common among the Zintl compounds mentioned above.

Methods

Polycrystalline samples of Eu5_5In2_2As6_6 have been previously synthesized [23, 24] but single crystals were not available until now. We report the first crystal growth of this compound here (Supplemental Material [22]). Powder x-ray diffraction was performed using a Bruker D8 ECO instrument. The FullProf suite [25] and VESTA software [26] were used for the Rietveld refinement and crystal visualization. Physical properties were measured using Quantum Design Dynacool-PPMS and MPMS-3.

Neutron diffraction data were collected using the thermal neutron triple-axis spectrometer VERITAS at the High Flux Isotope Reactor (HFIR) in Oak Ridge National Laboratory. A few crystals were mounted on an Al plate with either the cc-axis or bb-axis perpendicular to the scattering plane to probe the (H,K,0)(H,K,0) or (H,0,L)(H,0,L) Bragg peaks, respectively. Representation analysis was performed using SARAh [27]. The Cooper-Nathans formalism was used to calculate the resolution function of VERITAS [28], and the Wuesch-Prewitt algorithm [29] was used to calculate the angle-dependent neutron transmission of the sample. The statistical error for each intensity point corresponds to 1 standard deviation.

Results and Discussion

Structural Analysis

From a chemical standpoint, Eu5_5In2_2As6_6 is considered a Zintl compound, i.e. it is an intermetallic material that exhibits semiconducting behavior due to the charge balance between large cationic and anionic complexes that make up its lattice structure [30]. In this compound, the positive charge of [Eu5_5]10+^{10+} is balanced by the negative charges of [In2_2As4_4]6−^{6-} and [As2_2]4−^{4-} complexes [31].

Eu5_5In2_2As6_6 has an orthorhombic unit cell with three inequivalent Eu sites and a characteristic chain stacking of atoms as shown in Figure 1a. Details of the x-ray refinement are presented in the supplemental Fig. S1 and Tables S1 and S2 [22]. The crystal structure of Eu5_5In2_2As6_6 belongs to the same non-symorphic space group, PbamPbam (No. 55), as its sister compounds, Ba5_5In2_2Sb6_6 and Eu5_5In2_2Sb6_6, which are predicted to be topological hourglass and axion insulators, respectively [11, 32].

Image not extracted

Figure 1. (a) Orthorhombic unit cell of Eu5_5In2_2As6_6 in the space group PbamPbam. (b) Total, electronic, and lattice heat capacity shown in magenta, blue, and black colors. The lattice background is modeled assuming anharmonic Debye model [21, 22] with ΘD=196\Theta_{\mathrm{D}} = 196 K. Inset shows the electronic heat capacity (blue) and entropy (red). (c) AC magnetic susceptibility curves suggest AFM ordering with in-plane moments and alternating direction along the cc-axis. The hysteresis loop at T<TN2T < T_{\mathrm{N2}} in the inset suggests FM ordering along the aa-axis. (d) Three steps in the M(H)M(H) curve when H∥aH \parallel a suggest spin flops on different Eu sites. (e) The AFM spin structure determined by neutron diffraction in subsection III E at T<TN1T < T_{\mathrm{N1}}. The ordered moments are FM-aligned along the aa-axis and AFM-aligned along the cc-axis. The AFM-1 structure persists to zero temperature. (f) At T<TN2T < T_{\mathrm{N2}}, new magnetic domains with k2=(0,0,0)\mathbf{k}_2 = (0, 0, 0) emerge in separate regions of the sample. According to the analysis of panel (b), these domains (AFM-0) constitute less than 20% of the sample volume. The diagram on the right shows the phase separation scenario where the sample volume is represented by a rectangular prism. The 3 different prisms sketch the spatial variation of magnetic domains assuming the three zero-field magnetic phases that we observed. The gray color represents paramagnetic (para.) domains, while the blue and red domains respectively correspond to k1\mathbf{k}_1 (panel (e)) and k2\mathbf{k}_2 domains (panel (f)).

A recent density functional theory (DFT) calculation shows that while Eu5_5In2_2Sb6_6 and Eu5_5In2_2Bi6_6 could be topological, Eu5_5In2_2As6_6 is a trivial semiconductor [20]. We specifically chose this compound to demonstrate that a large CMR does not rely on topological aspects of the band structure. This is important because the salient features of CMR in the trivial semiconductor Eu5_5In2_2As6_6 are similar to those in topological systems such as the Dirac semimetals CeSbTe and EuAuSb [17, 18], Weyl semimetal EuCd2_2As2_2 [4], and axion insulator Eu5_5In2_2Sb6_6 [15].

Another exotic mechanism proposed for CMR is the modification of chiral orbital currents (COC) in Mn3_3Si2_2Te6_6 with applied field [16]. Central to this mechanism is the octahedral coordination of six Te atoms around each Mn2+^{2+} ion, enabling the interaction between the COC and Mn2+^{2+} spin. As seen in Figure 1a, the Eu2+^{2+} ions in Eu5_5In2_2As6_6 are not directly surrounded by a high-symmetry polyhedron of In and As. Therefore, COC is not a viable mechanism for the CMR in Eu5_5In2_2As6_6.

The above structural analysis dismisses topology and COC as the source of CMR in Eu5_5In2_2As6_6. Next, we will discuss its magnetic structure and phase diagram, before a detailed analysis of CMR.

Magnetic Transitions

The heat capacity data in Figure 1b reveal two AFM transitions in zero-field at TN1=16.0(2)T_{\mathrm{N1}} = 16.0(2) K and TN2=5.4(3)T_{\mathrm{N2}} = 5.4(3) K. Magnetism is driven by Eu2+^{2+} ions in the half-filled 4f74f^7 configuration (8S7/2{}^8S_{7/2}) with L=0L = 0 and S=7/2S = 7/2, consistent with the saturated entropy Rln⁡(8)R \ln(8) J mol−1^{-1} K−1^{-1} in the inset of Figure 1b, saturated magnetization 7.2 μB7.2\ \mu_{\mathrm{B}} in Figure 1d, and effective moment μeff=7.8(6) μB\mu_{\mathrm{eff}} = 7.8(6)\ \mu_{\mathrm{B}} from a Curie-Weiss analysis in Figure S2.

Image not extracted

Figure S2. Curie-Weiss analysis on the AC susceptibility data for both in-plane and out-of-plane field directions.

The inset of Figure 1b shows that approximately 10% of the full entropy is released by TN2T_{\mathrm{N2}}, 70% is released between TN2T_{\mathrm{N2}} and TN1T_{\mathrm{N1}}, and 20% is released above TN1T_{\mathrm{N1}}.

For comparison, Eu5_5In2_2Sb6_6 also exhibits two transitions at TN1=14T_{\mathrm{N1}} = 14 K and TN2=7T_{\mathrm{N2}} = 7 K but it releases nearly equal amounts of entropy at each transition [13]. In both materials, 20% of total entropy is released above TN1T_{\mathrm{N1}}, well into the paramagnetic (PM) phase, suggesting short-range magnetic correlations (polarons) that survive up to 3TN13T_{\mathrm{N1}} [33, 34].

Temperature dependence of the in-plane and out-of-plane magnetic susceptibility in Figure 1c suggests an AFM order with the moments in the abab plane but alternating along the cc-axis. The inset of Figure 1c shows the absence of a hysteresis loop at TN2<T<TN1T_{\mathrm{N2}} < T < T_{\mathrm{N1}}, so the order is purely AFM in this temperature range. However, a small hysteresis loop opens below TN2T_{\mathrm{N2}} when H∥aH \parallel a, suggesting a finite in-plane FM component when T<TN2T < T_{\mathrm{N2}}.

The behavior of M(H)M(H) curves along three different crystallographic directions is shown in Figure 1d. When H∥cH \parallel c, the M(H)M(H) curve is strictly linear, indicating an AFM order along the cc-axis. In contrast, the M(H)M(H) curve exhibits three steps when H∥aH \parallel a, indicating consecutive spin flops on different Eu sites within the abab-planes. The steps are less apparent when H∥bH \parallel b, suggesting that the aa-axis is the easy axis. A positive Curie-Weiss temperature (ΘCW=+15\Theta_{\mathrm{CW}} = +15 K) in Figure S2 indicates FM correlations despite an AFM order, consistent with FM correlations within the abab-planes in spite of AFM ordering along the cc-axis.

The spin structure at zero-field is discussed in subsection III E. While we cannot rule-out the possibility that both magnetic ordering at TN1T_{\mathrm{N1}} and TN2T_{\mathrm{N2}} form within the same magnetic domain, we suggest an alternative scenario where spatially separated magnetic domains form independently below each transition temperature. In this scenario, the majority domains (more than 80% of the sample volume according to the entropy release in Figure 1b) form below TN1T_{\mathrm{N1}} with a spin structure shown in Figure 1e. This structure is consistent with the χ(T)\chi(T) and M(H)M(H) curves shown in Figure 1c,Figure 1d. Specifically, it shows FM spins within the abab-planes with an easy aa-axis, and with alternating directions along the cc-axis. Less than 20% of the sample volume remains paramagnetic down to TN2T_{\mathrm{N2}} where it orders into the structure shown in Figure 1f, which has an FM component along the aa-axis and explains the hysteresis observed in the inset of Figure 1c at 2 K. Ferromagnetism along the aa-axis was also observed in the sister compound Eu5_5In2_2Sb6_6, but is only stabilized for TN2<T<TN1T_{\mathrm{N2}} < T < T_{\mathrm{N1}} [11, 13, 15].

Phase Diagrams

We used detailed measurements of the AC susceptibility and nanocalorimetry to map the temperature-field phase diagram of Eu5_5In2_2As6_6. Figure 2a shows three characteristic fields (H1H_1, H2H_2, and H3H_3) in the AC susceptibility data at T=2T = 2 K. They coincide with the three steps in the M(H∥a)M(H \parallel a) curve in Figure 1d. With increasing temperature, these characteristic fields become smaller and merge as shown in the phase diagram of Figure 2c. The triangular data points that define the outer boundary of the H∥aH \parallel a phase diagram indicate the field-dependence of TN1T_{\mathrm{N1}} in the heat capacity data (Figure 2b). The TN2T_{\mathrm{N2}} phase boundary is determined via a low-temperature nanocalorimetry technique [35] shown in Figure S3 (see also the supplemental tunnel diode oscillation data in Figure S4 and Figure S5). The striped region on the phase diagrams of Figure 2c indicates the phase separation between AFM-0 and AFM-1 spin structures in different regions of the sample with AFM-1 being the majority domain.

Image not extracted

Figure 2. (a) Field dependence of the real part of the AC susceptibility χ′(H)\chi'(H) at several temperatures. The H1H_1, H2H_2, and H3H_3 features imply spin flop transitions. They correspond to the three steps in DC magnetization in the red curve on Figure 1d. (b) Temperature dependence of the heat capacity at several magnetic fields. The sharp peak marks TN1T_{\mathrm{N1}}. (c) The magnetic phase diagram when H∥aH\parallel a. The area marked by striped lines indicates the coexistence of AFM-0 and AFM-1 domains (Figure 1e,Figure 1f) in separate regions of the sample. (d,e,f) Similar data and phase diagram to (a,b,c) when H∥cH\parallel c.

Image not extracted

Figure S3. Heat capacity data obtain by a nanocalorimeter from 20 to 0.05 K. Both TN1T_{\mathrm{N1}} and TN2T_{\mathrm{N2}} transitions are suppressed by an external magnetic field. These data were particularly helpful in mapping the TN2T_{\mathrm{N2}} phase boundary (Figure 2c,Figure 2f in the main text).

Image not extracted

Figure S4. Field dependency in TDO frequency at different temperatures.

Image not extracted

Figure S5. Phase diagram including the TDO data.

The AC susceptibility and heat capacity data for H∥cH \parallel c in Figure 2d,Figure 2e are used to map the phase diagram with out-of-plane field in Figure 2f. Supplemental nanocalorimetry data in Figure S3 are used to define the TN2T_{\mathrm{N2}} boundary when H∥cH \parallel c. The rich phase diagrams of Figure 2c,Figure 2f suggest that Eu5_5In2_2As6_6 is a multi-critical system [36] with several magnetic states accessible by modest magnetic fields. This is the result of having three inequivalent Eu sites in a non-symmorphic crystal structure (Figure 1a,Figure 1e,Figure 1f) that enables asymmetric exchange interactions [36].

Colossal Magnetoresistance (CMR)

The resistivity data reveal two types of CMR in Eu5_5In2_2As6_6, one at low temperatures near TN1T_{\mathrm{N1}} associated with a sharp upturn in the resistivity, and another at high temperatures well above TN1T_{\mathrm{N1}} associated with a broad peak in the resistivity. The resistivity upturn and resistivity peak are affected differently by the magnetic field, leading to two distinct types of CMR. Figure 3a shows the ρzz(T)\rho_{zz}(T) curves obtained with both current and voltage leads along the crystallographic cc-axis. The pink curve is measured at zero field, and the rest of the curves are obtained under different magnetic fields labeled by different colors in the legend of Figure 3b. For consistency, we follow the same color code to specify the magnetic field in Figure 3a,Figure 3b,Figure 3d,Figure 3e.

Image not extracted

Figure 3. (a) Resistivity curves measured with current and voltage leads along the cc-axis (ρzz\rho_{zz}) at different fields with H∥aH\parallel a. At zero field, ρzz\rho_{zz} exceeds the detection limit of our instrument below 20 K. The inset magnifies ρzz(T)\rho_{zz}(T) to indicate TpeakρT^\rho_{\mathrm{peak}}, TtroughρT^\rho_{\mathrm{trough}} and TupρT^\rho_{\mathrm{up}}. (b) ρzz(T)\rho_{zz}(T) measured with H∥cH\parallel c shows similar behavior as with H∥aH\parallel a. The inset highlights the different field dependence between the two types of CMR, namely the peak-type CMR (ρ(Tpeakρ)\rho(T^\rho_{\mathrm{peak}})) and upturn-type CMR (ρ(2 K)\rho(2\ \mathrm{K})). The ρzz(T)\rho_{zz}(T) data from which this plot is constructed are shown in Figure S10. (c) The broad peak in the AC susceptibility data at T>TN1T>T_{\mathrm{N1}} is evidence of polarons. The sharp peak is due to magnetic ordering. (d) Resistivity curves measured with current and field along the aa-axis (ρxx\rho_{xx}) with H∥bH\parallel b. The inset shows a crystal of Eu5_5In2_2As6_6 with the longest dimension being along the cc-axis. (e) ρxx(T)\rho_{xx}(T) curves with H∥cH\parallel c. The inset compares the field dependence of TpeakρT^\rho_{\mathrm{peak}}, TupρT^\rho_{\mathrm{up}}, and TN1T_{\mathrm{N1}}. (f) AC susceptibility data with H∥aH\parallel a shows similar behavior as with H∥cH\parallel c. The inset compares the field dependence of TpeakρT^\rho_{\mathrm{peak}} (circles), TpeakχT^\chi_{\mathrm{peak}} (squares), and TN1T_{\mathrm{N1}} (triangles).

The blue curve in Figure 3a shows that at 1.5 T, ρzz(T)\rho_{zz}(T) increases as the temperature is decreased from 300 K and forms a broad peak at around 27 K, but before it exhibits a sharp upturn at 15.5 K. Looking closely at the blue curve (inset of Figure 3a), one could identify a shallow trough near 24 K between the peak at 27 K and the upturn at 15.5 K. The shallow trough becomes deeper with increasing field but its characteristic temperature does not change. These three features in the resistivity curves, namely the broad peak at TpeakρT_{\mathrm{peak}}^\rho, the trough at TtroughρT_{\mathrm{trough}}^\rho, and the upturn at TupρT_{\mathrm{up}}^\rho, evolve with increasing magnetic field as shown in the inset of Figure 3e.

The suppression of the resistivity peak and resistivity upturn in Figure 3a leads to two distinct types of CMR as the field is increased from zero to 9 T. The high-temperature peak-type CMR at TpeakρT_{\mathrm{peak}}^\rho spans 3 orders of magnitude while its characteristic temperature TpeakρT_{\mathrm{peak}}^\rho increases with increasing field (inset of Figure 3e). The low-temperature upturn-type CMR is even larger, spanning 7 orders of magnitude. Changing the magnetic field direction from H∥aH\parallel a (Figure 3a) to H∥cH\parallel c (Figure 3b) does not affect either type of CMR. Changing the electric current direction also leaves both types of CMR qualitatively unchanged as seen in the ρxx(T)\rho_{xx}(T) curves with H∥bH\parallel b (Figure 3d) and H∥cH\parallel c (Figure 3e). Inset of Figure 3b reveals a gradual suppression of the resistivity peak at TpeakρT^\rho_{\mathrm{peak}} in contrast to the rapid drop in the resistivity upturn at 2 K. Note that the drop in the resistivity upturn is complete by 5.5 T, the field at which the AFM order is fully suppressed (Figure 2c,Figure 2f), whereas the suppression of resistivity peak extends to 9 T. Since the peak-type and upturn-type CMRs are marked by different resistivity features, different temperature scales, and different field dependencies, they must have different origins. Below, we will discuss each type in more detail.

curves at TpeakρT^\rho_{\mathrm{peak}} could be attributed to magnetic polarons, which are nanometer size FM clusters of Eu2+^{2+} moments coupled via conduction electrons [37, 38, 39, 40, 41]. When the sample is cooled from room temperature to TpeakρT^\rho_{\mathrm{peak}}, polarons proliferate and grow in size, so they scatter electrons at a higher rate and increase the resistivity. With further decreasing temperature below TpeakρT^\rho_{\mathrm{peak}}, polarons overlap and provide percolation paths for conduction electrons, so the resistivity decreases. Increasing the magnetic field makes it easier to polarize the FM clusters, so the polaron peak at TpeakρT^\rho_{\mathrm{peak}} shifts to higher temperatures with increasing field, as seen in Figure 3a,Figure 3b,Figure 3d,Figure 3e.

The inset of Figure 3e highlights the opposite effect of the magnetic field on magnetic polarons and magnetic order. The red circles mark TpeakρT^\rho_{\mathrm{peak}} in ρxx(T)\rho_{xx}(T) curves at different fields, while the green triangles mark TN1T_{\mathrm{N1}} from heat capacity data in Figure 2e. With increasing field, $T^\rho_{\mathrm{peak}}

Peak-type CMR. The broad peak in the resistivity is shifted to higher temperatures because local moments polarize more easily and fluctuate less under an external field. In contrast, TN1T_{\mathrm{N1}} is shifted to lower temperatures since the AFM order is suppressed by the field.

We also find evidence of magnetic polarons in the AC susceptibility data in Figure 3c and Figure 3f for H∥cH\parallel c and H∥aH\parallel a, respectively. Since Eu ff-moments are localized, a Curie-Weiss behavior (χ′∝1/T\chi^\prime\propto1/T) is expected at T>TN1T > T_{\mathrm{N1}}. However, Figure 3c and Figure 3f show a deviation from the 1/T1/T behavior with a broad peak at TpeakχT_{\mathrm{peak}}^{\chi} (inset of Figure 3c). The broad peak at TpeakχT_{\mathrm{peak}}^{\chi} suggests a reduction of the magnetic susceptibility due to short-range magnetic correlations developing within magnetic polarons.

In both susceptibility and resistivity data, the broad polaronic peaks are shifted to higher temperatures with increasing magnetic field, as shown in the inset of Figure 3f. However, the susceptibility peaks occur at lower temperatures compared to the resistivity peaks (Tpeakχ<TpeakρT_{\mathrm{peak}}^{\chi} < T_{\mathrm{peak}}^{\rho} in Figure 3f). This is because the susceptibility and resistivity reveal two different aspects of polarons. Susceptibility measures the spin correlations within polarons that develop at lower temperatures, while resistivity measures the scattering of conduction electrons by polarons that starts from higher temperatures. Such a difference between transport and thermodynamic quantities has also been reported in EuB6_6, the archetypal polaronic material [42, 43].

Upturn-type CMR. The second type of CMR is marked by the 7 orders-of-magnitude suppression of the resistivity upturn at the lowest measured temperature (2 K) as the field increases from zero to 9 T. The inset of Figure 3a shows that the resistivity upturn initiates at TtroughρT_{\mathrm{trough}}^{\rho} but then it accelerates by the onset of magnetic ordering at TupρT_{\mathrm{up}}^{\rho} which coincides with TN1T_{\mathrm{N1}} (inset of Figure 3e). At T<TupρT < T_{\mathrm{up}}^{\rho}, resistivity continues to increase down to the lowest measured temperatures. As the magnetic order becomes weaker with increasing field, the resistivity upturn also attenuates (Figure 3a,Figure 3b,Figure 3d,Figure 3e). Specifically, the inset of Figure 3b shows a near complete suppression of the resistivity upturn by 5.5 T, where the AFM order is fully suppressed. This is consistent with the parallel suppression of TupρT_{\mathrm{up}}^{\rho} and TN1T_{\mathrm{N1}} with increasing field from zero to 5.5 T in the inset of Figure 3e.

The parallel suppression of the upturn-type CMR and magnetic ordering with increasing field may suggest that the upturn-type CMR is caused by magnetic ordering. However, magnetic ordering typically reduces electrical resistance by removing spin fluctuations as a source of scattering. Thus, we can propose two potential mechanisms for the observed upturn-type CMR. The first possibility is a reconstruction of the electronic structure by the AFM order, i.e. a metal-insulator transition induced by the Slater mechanism [44]. The second possibility is a charge ordering concomitant with AFM ordering, a scenario which is well established in manganite perovskites with CMR [45, 46, 47]. The first scenario is challenged by DFT calculations that suggest Eu5_5In2_2As6_6 is a narrow gap semiconductor, so it already has a small gap

before entering the AFM phase [20]. The second scenario is challenged by the fact that unlike manganites, Eu5_5In2_2As6_6 is not a Mott insulator, so its physics could be different from those materials. Future ARPES experiments could reveal the band reconstruction [19, 48] while inelastic x-ray scattering [45] and microwave impedance microscopy [49] could reveal charge ordering.

It is worth noting that the temperature scale TtroughρT_{\mathrm{trough}}^{\rho}, where the resistivity upturn initiates, is field independent, as shown in the inset of Figure 3e. Notably, this feature in resistivity shows no corresponding thermodynamic signature either in magnetization or heat capacity measurements. Thus, the origin of this feature remains unclear and requires further investigation.

Neutron Diffraction

We used neutron diffraction data to resolve the spin structure of the ordered state below TN1T_{\mathrm{N1}} and TN2T_{\mathrm{N2}} in Eu5_5In2_2As6_6 at zero field. Figure 4a shows the onset of magnetic diffraction peaks indexed by a propagation vector k1=(0,0,12)\mathbf{k}_1=(0,0,\frac{1}{2}) below TN1T_{\mathrm{N1}} and another set of peaks indexed by k2=(0,0,0)\mathbf{k}_2=(0,0,0) below TN2T_{\mathrm{N2}}. Given the likelihood of a charge ordering in Eu5_5In2_2As6_6, we associated this phenomenology to a magnetic phase segregation scenario where k1\mathbf{k}_1 and k2\mathbf{k}_2 magnetic domains form separately within the sample. This is consistent with the lack of intensity changes observed at Q=(0,0,12)\mathbf{Q}=(0,0,\frac{1}{2}) when the $\mathbf{k}_2 peaks onsets below TN2T_{\mathrm{N2}}. The k1\mathbf{k}_1 wave vector corresponds to an inter-layer AFM order along the cc-axis, which persists to zero temperature, while k2\mathbf{k}_2 corresponds to the development of a spin structure with an FM inter-layer coupling arising below TN2T_{\mathrm{N2}}. Note that the peak intensity of k2\mathbf{k}_2 is about 10 times smaller than k1\mathbf{k}_1, indicating that the k2\mathbf{k}_2 domains form the majority of the sample volume. For comparison, Eu5_5In2_2Sb6_6 shows a reverse order of wave vectors, such that k2\mathbf{k}_2 and k1\mathbf{k}_1 appear with comparable intensities below TN1T_{\mathrm{N1}} and TN2T_{\mathrm{N2}}, respectively [13].

Image not extracted

Figure 4. (a) Temperature dependence of the neutron magnetic Bragg peak intensities at Q=(0,0,12)\mathbf{Q}=(0,0,\frac{1}{2}) and Q=(1,0,0)\mathbf{Q}=(1,0,0) in the top panel is compared to the temperature dependence of heat capacity in the bottom panel. (b) The top panel shows the intensity of Q=(6,0,32)\mathbf{Q}=(6,0,\frac{3}{2}) peaks at both T=1.5T=1.5 K and T=20T=20 K. The bottom panel shows that magnetic Bragg scattering is present for Q=(1,0,0)\mathbf{Q}=(1,0,0) (triangles) but absent for Q=(0,1,0)\mathbf{Q}=(0,1,0) (squares) at T=1.5T=1.5 K.

Following the symmetry analysis of Refs. [13, 15] for Eu5_5In2_2Sb6_6, we find 8 irreducible representations (irreps) for the Eu5_5In2_2As6_6 space group PbamP bam compatible with the propagation vectors k1=(0,0,12)\mathbf{k}_1 = (0, 0, \frac{1}{2}) and k2=(0,0,0)\mathbf{k}_2 = (0, 0, 0). However, only 4 irreps (Γ1\Gamma_1, Γ3\Gamma_3, Γ5\Gamma_5, Γ7\Gamma_7) can produce magnetization on all three Eu Wyckoff sites – a condition imposed by the full entropy of Rln⁡8R \ln8 in Figure 1b. These four irreps correspond to an AFM or FM arrangement of Eu2+^{2+} spins within each abab-plane oriented along either the aa (Γ3\Gamma_3, Γ5\Gamma_5), bb (Γ3\Gamma_3, Γ5\Gamma_5), or cc (Γ1\Gamma_1, Γ7\Gamma_7) directions. The second-order phase transitions revealed by heat capacity (Figure 1b) and order parameter measurements (Figure 4a) require a single irrep to explain the magnetic structure.

We first analyze the k1\mathbf{k}_1 magnetic domains that onset below TN1T_{\mathrm{N1}} and persist to zero temperature with at least 70% volume fraction according to the entropy analysis in Figure 1b. As determined by the order parameter curve collected at Q=(0,0,12)\mathbf{Q} = (0, 0, \frac{1}{2}) (Figure 4a), magnetic Bragg peaks for these domains are observed at Q=(0,0,L)\mathbf{Q} = (0, 0, L). Since neutron scattering only probes magnetization perpendicular to the momentum transfer Q\mathbf{Q}, any irrep with spins strictly along the cc-direction (Γ1\Gamma_1 and Γ7\Gamma_7) are excluded. This leaves Γ3\Gamma_3 and Γ5\Gamma_5 as the only viable choices.

Γ3\Gamma_3 has an FM basis vector with the spins pointing along the aa-axis (Γ3(ψ1)\Gamma_3(\psi_1)), and an AFM basis vector with the spins pointing along the bb-axis (Γ3(ψ2)\Gamma_3(\psi_2)). Γ5\Gamma_5 has the opposite spin arrangement, i.e. it has an FM basis vector with spins along the bb-axis (Γ5(ψ2)\Gamma_5(\psi_2)), and an AFM basis vector with spins along the aa-axis (Γ5(ψ1)\Gamma_5(\psi_1)). An FM order within the abab plane produces magnetic Bragg peaks at Q=(even,0,L2)\mathbf{Q} = (\text{even}, 0, \frac{L}{2}), while an AFM order produces them at Q=(odd,0,L2)\mathbf{Q} = (\text{odd}, 0, \frac{L}{2}). The top panel of Figure 4b shows magnetic Bragg scattering at even positions, indicating an FM arrangement within the abab-plane. Therefore, the spin structure can be represented by either Γ3(ψ1)\Gamma_3(\psi_1) or Γ5(ψ2)\Gamma_5(\psi_2). A quantitative refinement of the Bragg peak intensities could distinguish between Γ3(ψ1)\Gamma_3(\psi_1) and Γ5(ψ2)\Gamma_5(\psi_2), but this contrast disappears due to the large neutron absorption cross-sections of Eu. However, the Γ3(ψ1)\Gamma_3(\psi_1) basis vector is consistent with the low-temperature magnetization of Eu5_5In2_2As6_6 (Figure 1d), which shows spin-flop transitions for field applied along the aa-axis. This contrasts with the magnetization for a field applied along the bb-axis where such spin-flop transitions are not observed and would be expected to occur for the Γ5(ψ2)\Gamma_5(\psi_2) structure. Therefore, we conclude that Γ3(ψ1)\Gamma_3(\psi_1) is the appropriate irrep to describe the magnetic domains forming below TN1T_{\mathrm{N1}} in Eu5_5In2_2As6_6. This irrep also describes the low-temperature magnetic phase of Eu5_5In2_2Sb6_6 [13, 15]. We depict the resulting k1\mathbf{k}_1 magnetic domains in Figure 1e.

We next turn our attention to the details of the k2\mathbf{k}_2 magnetic domains. Our bulk magnetization data in Figure 1c shows a hysteresis loop opening below TN2T_{\mathrm{N2}}, indicating a net FM component along the aa-axis. We also detected magnetic Bragg peaks at structurally non-allowed Bragg positions, such as Q=(1,0,0)\mathbf{Q} = (1,0,0) plotted in the bottom panel of Figure 4b. This indicates that the k2=(0,0,0)\mathbf{k}_2 = (0,0,0) domains also have an AFM component within the abab plane. Since no magnetic Bragg peaks were found at Q=(0,H,0)\mathbf{Q} = (0,H,0) positions (Figure 4b, bottom panel), the spin polarization of this AFM component is along the bb-axis. Thus, the magnetic structure below TN2T_{\mathrm{N2}} must have a net FM component along the aa-axis and an AFM component along the bb-axis. This means the irrep for the k2=(0,0,0)\mathbf{k}_2 = (0,0,0) domains are identical to the ones for k1=(0,0,12)\mathbf{k}_1 = (0,0,\frac{1}{2}), but the spins do not reverse going along the cc-direction [13, 15]. The spin structure for the k2\mathbf{k}_2 domains is represented in Figure 1f.

Finally, we note that our neutron diffraction data cannot distinguish a multi-k\mathbf{k} structure from our proposed magnetic phase segregation scenario as they both lead to identical zero-field neutron diffraction patterns. Magnetic phase segregation between the same AFM-1 and AFM-0 domains was also suggested for Eu5_5In2_2Sb6_6 [15], but in this latter compound the AFM-0 state forms the majority domains. Furthermore, we propose that a phase separation is quite likely if charge ordering/separation is responsible for the upturn-type CMR. The competition between charge order and AFM order over the same carriers leads to enhanced scattering of conduction electrons, giving rise to the resistivity upturn. With increasing field, both charge and AFM order are suppressed, giving rise to the upturn-type CMR. Future studies using magnetic imaging techniques may reveal magnetic and charge phase separation in Eu5_5In2_2As6_6.

For completeness, we also discuss the multi-k\mathbf{k} scenario in the supplemental information [22].

Summary and Outlook

The two types of CMR discussed here have been observed in different Zintl compounds but they were not explained in a unifying manner. For example, CeSb0.1_{0.1}Te1.9_{1.9}, EuAuSb, and EuCd2_2As2_2 show the peak-type CMR [4, 17, 18] whereas EuCd2_2P2_2, Mn3_3Si2_2Te6_6 and Eu5_5In2_2Sb6_6 show both the peak-type CMR and the upturn-type CMR [8, 11, 16]. Based on the similarity of transport data across different Zintl compounds, we argue that similar mechanisms are at work to produce the two types of CMR in these materials. Specifically, we attribute the peak-type CMR to percolation of magnetic polarons and the upturn-type CMR to either melting of a charge ordered state or band structure reconstruction by the AFM order (Slater mechanism).

The CMR induced by magnetic polarons implies a coupling between short-range FM ordering and charge transport, while the CMR associated with charge ordering reflects a coupling between long-range AFM ordering and charge transport. Such an intricate interplay between magnetism (both long-range and short-range) and the charge transport is a key feature of many Eu-based Zintl phases [9, 15, 50], potentially driven by the large spin of Eu2+^{2+} and considerable magnetoelastic coupling. It is noteworthy that a strong influence of the 4f4f states of Europium, through their coupling to valence and conduction states, on the band structure and charge transport of EuCd2_2X2_2 (XX = P,As,Sb) systems has already been demonstrated [48, 51].

An indirect manifestation of magnetoelastic coupling is the dependence of the magnetic state of Eu-based Zintl compounds on the choice of flux in crystal growth [5, 52]. These characteristics position Eu-based CMR materials as promising candidates for magnetic sensing, spin valve, and piezoresistive technologies [12].

In the supplemental Figure S6 and Figure S7, we present an Arrhenius analysis on the zero-field resistivity data from two samples of Eu5_5In2_2As6_6 revealing significantly different activation gaps of 35 meV and 125 meV in samples S1 and S2, respectively. Hall effect data in Figure S8 reveal different n-type carrier concentrations in samples S1 (101810^{18} cm−3^{-3}) and S2 (101610^{16} cm−3^{-3}), so the sample with a larger gap (S1) shows a smaller carrier concentration. The different gap values and carrier concentrations in different samples suggest that Eu5_5In2_2As6_6 is a self-doped semiconductor. Remarkably, all the interesting transport phenomena in this material, and possibly other Eu-based CMR materials, seem to originate from a small density of extrinsic carriers.

Image not extracted

Figure S7. Arrhenius analysis on the resistivity data from samples S1 and S2.

Image not extracted

Figure S6. Sample dependence of the high-temperature resistivity.

Image not extracted

Figure S8. Hall resistance as function of field in samples S1 and S3 at room temperature.

Acknowledgments

F.T. and X.Y. acknowledge funding from the U.S. Department of Energy under Award No. SC0023124. This material is based upon work supported by the Air Force Office of Scientific Research under award numbers FA2386-21-1-4059 and FA9550-23-1-0124. A. R. acknowledges support from the Swedish Research Council, D. Nr. 2021-04360. A portion of this work was performed at the National High Magnetic Field Laboratory, which is supported by the National Science Foundation Cooperative Agreement No. DMR-1644779 and the state of Florida. The support for neutron scattering was provided by the Center for High-Resolution Neutron Scattering, a partnership between the National Institute of Standards and Technology and the National Science Foundation under Agreement No. DMR-2010792. The identification of any commercial product or trade name does not imply endorsement or recommendation by the National Institute of Standards and Technology. A portion of this research used resources at the High Flux Isotope Reactor, a DOE Office of Science User Facility operated by the Oak Ridge National Laboratory. The beam time was allocated to VERITAS on proposal number IPTS-31509.

Crystal Growth and X-Ray Diffraction

Large crystals of Eu5_5In2_2As6_6 were grown using the flux method. The crystals are shiny and pillar-shaped with the longest dimension along the cc-axis as shown in the main text Figure 1. They are stable in air but undergo hydrolysis when exposed to moisture, releasing arsine gas and leaving a red powder. It is advisable to use a fume hood while polishing the samples to avoid potential arsine poisoning.

Powder x-ray diffraction (PXRD) was used to determine the crystal structure of Eu5_5In2_2As6_6. The Rietveld refinement is shown in Figure S1, where the observed Bragg peaks appear in red and the calculated pattern in blue. The calculated intensities correspond to an orthorhombic unit cell in the space group PbamPbam with three Eu, one In, and three As sites as summarized in Table S1. The structural parameters and fit quality factors are summarized in Table S2.

Image not extracted

Figure S1. Rietveld refinement of the PXRD pattern of Eu5_5In2_2As6_6 in space group PbamPbam (No. 55). The red, blue, and green curves correspond to the observed, calculated, and difference intensities, respectively, while the black ticks mark the expected positions of the Bragg peaks.

atomsitexxyyzzocc.BisoB_{\mathrm{iso}} (Å2^2)
Eu14g0.32500.00000.000000.5001.0
Eu24g0.08530.25290.000000.5001.0
Eu32a0.00000.00000.000000.2501.0
In14h0.32910.21600.500000.5001.0
As14g0.33640.32470.000000.5001.5
As24h0.15320.09870.500000.5001.5
As34h0.02080.41050.500000.5001.5

Table S1. Atomic coordinates, site occupancies, and isotropic Debye-Waller factors from PXRD Rietveld refinement of Eu5_5In2_2As6_6 in space group PbamPbam at room temperature.

Unit cell parametersRefinement parameters
Space GroupPbamPbamParameters20
aa (Å)11.8839(5)RBraggR_{\mathrm{Bragg}} (%)15.0
bb (Å)13.7885(6)RFR_{\mathrm{F}} (%)18.4
cc (Å)4.3506(2)RexpR_{\mathrm{exp}} (%)29.1
VV (Å3^3)712.909RpR_{\mathrm{p}} (%)37.8
ZZ2RwpR_{\mathrm{wp}} (%)33.1
ρ\rho (gr cm−3^{-3})6.703χ2\chi^21.30

Table S2. Unit cell parameters of Eu5_5In2_2As6_6 and quality factors of the PXRD Rietveld refinement at room temperature.

Curie-Weiss Analysis

The Curie-Weiss (CW) analysis is shown in Figure S2 for both H∥aH\parallel a (red) and H∥cH\parallel c (blue) directions. Note that the high-temperature tail of the magnetic susceptibility is nearly identical for both field directions, so the same fit goes through both data sets. It yields the Curie-Weiss temperature ΘCW=15.0(2)\Theta_{\mathrm{CW}} = 15.0(2) K and effective moment μeff=7.8(6) μB\mu_{\mathrm{eff}} = 7.8(6)\,\mu_{\mathrm{B}}. Despite the AFM order at TN1=16.0(2)T_{\mathrm{N1}} = 16.0(2) K, the CW temperature is positive due to the in-plane FM correlations among Eu2+^{2+} moments (see Figure 1e in the main text). The absence of a splitting between the field-cooled (FC) and zero-field-cooled (ZFC) data, shown respectively by full and empty symbols in

Figure S2, confirms AFM ordering.

Heat Capacity

In the main text, the anharmonic Debye model was used to subtract the lattice contribution to the heat capacity. Before using such a model, several attempts were made to grow crystals of Sr5_5In2_2As6_6 and Ba5_5In2_2As6_6, which could have been used as a lattice model for the title compound. These attempts were unsuccessful, hence a model

calculation was used for the phonon background subtraction. The simplest expression for the heat capacity in the anharmonic Debye model involving only nearest neighbor interactions in a body-centered cubic lattice is

C=9R(TΘD)3∫0ΘD/Tx4ex dx(ex−1)2(1+Ax)(1)C = 9R\left(\frac{T}{\Theta_{\mathrm{D}}}\right)^3 \int_{0}^{\Theta_{\mathrm{D}/T} \frac{x^4 e^x\,\mathrm{d}x}{(e^x-1)^2(1+Ax)}} \tag*{(1)}

where AA is a constant and x=ℏck/kBTx = \hbar c k/k_{\mathrm{B}}T (Stern et al. [29]). The black dashed line in the main Figure 1b represents the heat capacity calculated by equation 1.

All the “raw” specific heat data used for mapping the phase diagram of Eu5_5In2_2As6_6 were shown in the main text Figure 2, except for a subset of data that was necessary to map the AFM-0 phase. It was necessary to use a nanocalorimeter at millikelvin temperatures to obtain the data in Figure S3

Tunnel Diode Oscillator (TDO)

We used TDO to confirm the boundaries of the phase-separated region, AFM-1 + 0, in the phase diagram of Eu5_5In2_2As6_6. Figure S4 shows the changes in the TDO frequency at different temperatures. By observing the changes in the TDO signal we can map, to a good extent, the H3H_3 and TN2T_{\mathrm{N2}} components of the phase diagram as shown in Figure S5. The red down-triangles from the TDO measurements confirm the boundaries of the AFM-1+0 phase determined from heat capacity measurements (green down-triangles).

Tunnel diode oscillator (TDO) measurements were performed with the sample coil being a part of an oscillator circuit at a resonant frequency of 65 MHz. The TDO circuit board and inductor containing the sample were cooled to low temperatures while a heterodyne circuit at room temperature amplified and mixed the RF signal for recording with a frequency counter.

Sample Dependence of the Resistivity and Hall Effect

Figure S6 shows that the high-temperature resistivity ρ(T)\rho(T) of different Eu5_5In2_2As6_6 samples could have different activation gaps. We extracted the gap values in samples S1 and S2 using an Arrhenius analysis as shown in Figure S7. We also measured Hall effect on both samples (Figure S8) and found a smaller carrier density in sample S2 that has a larger gap. The carrier concentration was

extracted using the single band relationship RH=1/neR_H = 1/ne where RHR_H, nn, and ee are the Hall coefficient, carrier concentration, and electron charge. From here, we estimate a concentration of 1.6×10181.6 \times10^{18} cm−3^{-3} and 8.4×10168.4 \times10^{16} cm−3^{-3} for n-type carriers in samples S1 and S2, respectively.

Such variations in the activation gap and carrier concentration in different samples suggest that Eu5_5In2_2As6_6 is a self-doped semiconductor with extrinsic carriers. These carriers are coming from the slight off-stoichiometry of Eu and As, confirmed by energy dispersive x-ray spectroscopy (EDX). The EDX measurements on 6 spots in 2 samples gave the average composition of Eu4.8_{4.8}In1.9_{1.9}As6.3_{6.3} with approximately 10% error for each element. The measurements were performed in an FEI Scios DualBeam electron microscope equipped with Oxford detector.

Spin Fluctuation Scaling Analysis

In this work, we argue that mangetic polarons are the source of CMR at high temperatures (T>TN1T > T_{\mathrm{N1}}). Polarons are coherent islands of FM spins polarized by coupling to itinerant electrons. They vary in size from a few to tens of angstroms. While polarons could be the source of the resistivity peak and the associated CMR, critical magnetic fluctuations could also produce a similar phenomenon.

Theoretically, it is expected that the magnetoresistance scales with magnetization when critical fluctuations are responsible for the CMR. Such a scaling analysis comes from a Ginzburg-Landau theory, which is explained in Ref. [40] and references thereof. In Figure S9, we show the scaling of low-field magnetoresistance with the square of the normalized magnetization, Δρρ0\frac{\Delta\rho}{\rho_0} (denoted as ρMR\rho_{\mathrm{MR}}) =C(MMsat)2= C\left(\frac{M}{M_{\mathrm{sat}}}\right)^2, where MsatM_{\mathrm{sat}} is the saturation magnetization and CC is the scaling factor. The scaling results in a value of C=150C = 150, which is two orders of magnitude larger than the expected value for critical fluctuations [40]. This suggests the mechanism for magnetoresistance is not due to the suppression of magnetic fluctuations. A similar deviation in the scaling constant is observed in EuB6_6 [40] and Eu5_5In2_2Sb6_6 [11] where percolation of polarons, as their size increases in the magnetic field, is argued to cause the magnetoresistance rather than the suppression of magnetic fluctuations.

Image not extracted

Figure S9. Magnetoresistance vs normalized magnetization squared.

Image not extracted

Figure S11. Fits to the variable range hopping model.

Additional Resistivity Data

In the inset of Figure 3b in the main text, we show ρ(2K)\rho(2\mathrm{K}) and ρ(Tpeak)\rho(T_{\mathrm{peak}}) to distinguish the two types of CMR based on how the resistivity peak and the resistivity upturn respond to the magnetic field. The resistivity data we used to make that plot is presented in Figure S10. These data are obtained from a different sample than the one we used in the main text. The salient features of resistivity are the same between these two different samples.

Image not extracted

Figure S10. High field ρzz\rho_{zz} curves showing the difference in field suppression of resistivity from two types of CMR. The inset highlights the 6–9 T curves, illustrating the difference between the two types of CMR.

Figure S11 shows fits to the variable range hopping model, confirming hopping-like transport of impurity states in the sample. Note that high-temperature resistivity fits a conventional semiconducting behavior, as shown in Figure S7, but the low-temperature data would not fit a single exponential. Instead, it follows a variable range hopping behavior. This analysis confirms that the conduction electrons in Eu5_5In2_2As6_6 come from impurity states with a hopping-type transport through different domains in the sample. Remarkably, it seems such extrinsic carriers from impurity states are responsible for the sensitive response of the title compound to the magnetic field.

Multi-kk Scenario

In the main text, we interpreted the neutron diffraction results in terms of phase separation between the AFM-1 and AFM-0 spin structures shown in Figure 1(e,Figure 1f). A different interpretation would be to evoke a multi-kk order below TN2T_{\mathrm{N2}}. In this case, a multi-kk structure implies both the k1=(0,0,12)\mathbf{k}_1 = (0, 0, \frac{1}{2}) and k2=(0,0,0)\mathbf{k}_2 = (0, 0, 0) form within the same domain. As reported in the main text, both k1\mathbf{k}_1 and k2\mathbf{k}_2 order within the Γ3\Gamma_3 irreducible representation, which both have a finite FM spin component polarized along the aa-axis. However, the Γ3(ψ1)\Gamma_3(\psi_1) for the k1=(0,0,12)\mathbf{k}_1 = (0, 0, \frac{1}{2}) order reverses its polarization along the cc-axis whereas the Γ3(ψ1)\Gamma_3(\psi_1) for the k2=(0,0,0)\mathbf{k}_2 = (0, 0, 0) order maintains its polarization along the cc-axis. If k1\mathbf{k}_1 and k2\mathbf{k}_2 coexist on the same Eu Wyckoff site, then it would lead to a spatial variation of the magnetization on such a site, which is nonphysical for a localized spin system. Assuming a multi-kk scenario, one needs to conclude that the k1\mathbf{k}_1 and k2\mathbf{k}_2 orders arise on different Eu sites. Considering the small entropy released at TN2T_{\mathrm{N2}} and the weak intensity of the k2\mathbf{k}_2 Bragg peaks, we postulate that the Eu3 sites, which have half the occupancy of the Eu1 and Eu2 sites, are the ones ordering with the k2\mathbf{k}_2 wave vector. Assuming homogeneous total magnetization on all three Eu sites, the spin structure of Eu5_5In2_2As6_6 is illustrated in Figure S12 for TN2<T<TN1T_{\mathrm{N2}} < T < T_{\mathrm{N1}} and in Figure S12 for T<TN2T < T_{\mathrm{N2}}. These spin structures are consistent with the χ(T)\chi(T) and M(H)M(H) curves shown in Figure 1c,Figure 1d in the main manuscript. Specifically, they show FM spins within the abab-plans with an easy aa-axis, and with alternating directions along the cc-axis. The Eu3 spins order only below TN2T_{\mathrm{N2}} in a way that amounts to a FM component along the aa-axis, explaining the hysteresis observed in the inset of Figure 1(c) at 2 K.

Image not extracted

Figure S12. The magnetic structure at (a) TN2<T<TN1T_{\mathrm{N2}} < T < T_{\mathrm{N1}} and (b) T<TN2T < T_{\mathrm{N2}} in a scenario where the magnetic order does not phase segregate, instead, the entire sample forms a multi-kk spin structure below TN2T_{\mathrm{N2}}.

Given the complexity of the multi-kk spin structure shown in Figure S12 and the likelihood of a charge segregation as the mechanism of resistivity upturn, we favor the magnetic phase separation scenario as presented in the main text.

References

  1. [1]V. Baltz, A. Manchon, M. Tsoi, T. Moriyama, T. Ono, and Y. Tserkovnyak, Antiferromagnetic spintronics, Reviews of Modern Physics 90, 015005 (2018).DOI
  2. [2]J. Han, R. Cheng, L. Liu, H. Ohno, and S. Fukami, Coherent antiferromagnetic spintronics, Nature Materials 22, 684 (2023).DOI
  3. [3]J. Xu, J. He, J.-S. Zhou, D. Qu, S.-Y. Huang, and C. Chien, Observation of Vector Spin Seebeck Effect in a Noncollinear Antiferromagnet, Physical Review Letters 129, 117202 (2022).DOI
  4. [4]J.-Z. Ma, S. M. Nie, C. J. Yi, J. Jandke, T. Shang, M. Y. Yao, M. Naamneh, L. Q. Yan, Y. Sun, A. Chikina, V. N. Strocov, M. Medarde, M. Song, Y.-M. Xiong, G. Xu, W. Wulfhekel, J. Mesot, M. Reticcioli, C. Franchini, C. Mudry, M. Müller, Y. G. Shi, T. Qian, H. Ding, and M. Shi, Spin fluctuation induced Weyl semimetal state in the paramagnetic phase of EuCd2As2, Science Advances 5, eaaw4718 (2019).
  5. [5]N. H. Jo, B. K. Kuthanazhi, Y. Wu, E. Timmons, T.-H. Kim, L. Zhou, L.-L. Wang, B. G. Ueland, A. Palasyuk, D. H. Ryan, R. J. McQueeney, K. Lee, B. Schrunk, A. A. Burkov, R. Prozorov, S. J. Bud’ko, A. Kaminski, and P. C. Canfield, Manipulating magnetism in the topological semimetal EuCd2As2, Physical Review B 101, 140402 (2020).
  6. [6]Y. Xu, Z. Song, Z. Wang, H. Weng, and X. Dai, Higher-Order Topology of the Axion Insulator EuIn2As2, Physical Review Letters 122, 256402 (2019).
  7. [7]S. X. M. Riberolles, T. V. Trevisan, K. Kuthanazhi, T. W. Heitmann, F. Ye, D. C. Johnston, S. L. Bud’ko, D. H. Ryan, P. C. Canfield, A. Kreyssig, A. Vishwanath, R. J. McQueeney, L.-L. Wang, P. P. Orth, and B. G. Ueland, Magnetic, crystalline-symmetry-protected axion electrodynamics and field-tunable unpinned Dirac cones in EuIn2As2, Nature Communications 12, 999 (2021).DOI
  8. [8]Z.-C. Wang, J. D. Rogers, X. Yao, R. Nichols, K. Atay, B. Xu, J. Franklin, I. Sochnikov, P. J. Ryan, D. Haskel, and F. Tafti, Colossal Magnetoresistance without Mixed Valence in a Layered Phosphide Crystal, Advanced Materials 33, 2005755 (2021).DOI
  9. [9]V. Sunko, Y. Sun, M. Vranas, C. C. Homes, C. Lee, E. Donoway, Z.-C. Wang, S. Balguri, M. B. Mahendru, A. Ruiz, B. Gunn, R. Basak, S. Blanco-Canosa, E. Schierle, E. Weschke, F. Tafti, A. Frano, and J. Orenstein, Spin-carrier coupling induced ferromagnetism and giant resistivity peak in EuCd2P2, Physical Review B 107, 144404 (2023).DOI
  10. [10]C. C. Homes, Z.-C. Wang, K. Fruhling, and F. Tafti, Optical properties and carrier localization in the layered phosphide EuCd2P2, Physical Review B 107, 045106 (2023).DOI
  11. [11]P. Rosa, Y. Xu, M. Rahn, J. Souza, S. Kushwaha, L. Veiga, A. Bombardi, S. Thomas, M. Janoschek, E. Bauer, M. Chan, Z. Wang, J. Thompson, N. Harrison, P. Pagliuso, A. Bernevig, and F. Ronning, Colossal magnetoresistance in a nonsymmorphic antiferromagnetic insulator, npj Quantum Materials 5, 1 (2020).
  12. [12]S. Ghosh, C. Lane, F. Ronning, E. D. Bauer, J. D. Thompson, J.-X. Zhu, P. F. S. Rosa, and S. M. Thomas, Colossal piezoresistance in narrow-gap Eu5In2Sb6, Physical Review B 106, 045110 (2022).DOI
  13. [13]V. C. Morano, J. Gaudet, N. Varnava, T. Berry, T. Halloran, C. J. Lygouras, X. Wang, C. M. Hoffman, G. Xu, J. W. Lynn, T. M. McQueen, D. Vanderbilt, and C. L. Broholm, Noncollinear 2k antiferromagnetism in the Zintl semiconductor Eu5In2Sb6, Physical Review B 109, 014432 (2024).DOI
  14. [14]J. C. Souza, S. M. Thomas, E. D. Bauer, J. D. Thompson, F. Ronning, P. G. Pagliuso, and P. F. S. Rosa, Microscopic probe of magnetic polarons in antiferromagnetic Eu5In2As6, Physical Review B 105, 035135 (2022).DOI
  15. [15]M. C. Rahn, M. N. Wilson, T. J. Hickin, F. L. Pratt, C. Wang, F. Orlandi, D. D. Khalyavin, P. Manuel, L. S. I. Veiga, A. Bombardi, S. Francoual, P. Bereciartu, A. S. Sukhano, J. D. Thompson, S. M. Thomas, P. F. S. Rosa, T. Lancaster, F. Ronning, and M. Janoschek, Unusual magnetism of the axion-insulator candidate Eu5In2Sb6 (2023).DOI
  16. [16]Y. Zhang, Y. Ni, H. Zhao, S. Hakani, F. Ye, L. DeLong, I. Kimchi, and G. Cao, Control of chiral orbital currents in a colossal magnetoresistance material, Nature 611, 467 (2022).DOI
  17. [17]R. Singha, K. J. Dalgaard, D. Marchenko, M. Krivenkov, E. D. L. Rienks, M. Jovanovic, S. M. L. Teicher, J. Hu, T. H. Salters, J. Lin, A. Varykhalov, N. P. Ong, and L. M. Schoop, Colossal magnetoresistance in the multiple wave vector charge density wave regime of an antiferromagnetic Dirac semimetal, Science Advances 9, eadh0145 (2023).DOI
  18. [18]D. Ram, J. Singh, S. Banerjee, A. Sundaresan, D. Samal, V. Kanchana, and Z. Hossain, Magnetotransport and electronic structure of EuAuSb: A candidate antiferromagnetic Dirac semimetal, Physical Review B 109, 155152 (2024).DOI
  19. [19]H. Zhang, F. Du, X. Zheng, S. Luo, Y. Wu, H. Zheng, S. Cui, Z. Sun, Z. Liu, D. Shen, M. Snidman, Y. Song, M. Shi, Z. Zhong, C. Cao, H. Yuan, and Y. Liu, Electronic band reconstruction across the insulator-metal transition in colossally magnetoresistive EuCd2P2, Physical Review B 108, L241115 (2023).DOI
  20. [20]N. Varnava, T. Berry, T. M. McQueen, and D. Vanderbilt, Engineering magnetic topological insulators in Eu5−xMxX6 Zintl compounds, Physical Review B 105, 235128 (2022).
  21. [21]E. A. Stern, Theory of the Anharmonic Properties of Solids, Physical Review 111, 786 (1958).DOI
  22. [22]See the Supplemental Material for additional information about crystal growth, Curie-Weiss analysis, low-temperature specific heat, tunnel diode oscillations, and transport data.
  23. [23]N. Tomitaka, Y. Goto, K. Morino, K. Hoshi, Y. Nakahira, H. Ito, A. Miura, H. Usui, and Y. Mizuguchi, Bipolar doping and thermoelectric properties of Zintl arsenide Eu5In2As6, Journal of Materials Chemistry A 9, 26362 (2021).
  24. [24]M. Radzieowski, F. Stegemann, S. Klenner, Y. Zhang, B. P. T. Fokwa, and O. Janka, On the divalent character of the Eu atoms in the ternary Zintl phases Eu5In2Pn6 and Eu2MAS3 (Pn = As–Bi; M = Al, Ga), Materials Chemistry Frontiers 4, 1231 (2020).
  25. [25]J. Rodríguez-Carvajal, Recent advances in magnetic structure determination by neutron powder diffraction, Physica B: Condensed Matter 192, 55 (1993).
  26. [26]K. Momma and F. Izumi, VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data, Journal of Applied Crystallography 44, 1272 (2011).DOI
  27. [27]A. S. Wills, A new protocol for the determination of magnetic structures using simulated annealing and representational analysis (SARAh), Physica B: Condensed Matter 276-278, 680 (2000).DOI
  28. [28]N. J. Chesser and J. D. Axe, Derivation and experimental verification of the normalized resolution function for inelastic neutron scattering, Acta Crystallographica Section A: Crystal Physics, Diffraction, Theoretical and General Crystallography 29, 160 (1973).DOI
  29. [29]B. J. Wuensch and C. T. Prewitt, Corrections for x-ray absorption by a crystal of arbitrary shape, Zeitschrift für Kristallographie - Crystalline Materials 122, 24 (1965).DOI
  30. [30]S. M. Kauzlarich, Zintl Phases: From Curiosities to Impactful Materials, Chemistry of Materials 35, 7355 (2023).DOI
  31. [31]A. B. Childs, S. Baranets, and S. Bobev, Five new ternary indium-arsenides discovered. Synthesis and structural characterization of the Zintl phases Sr3In2As4, Ba3In2As4, Eu3In2As4, SrIn2As6 and Eu5In2As6, Journal of Solid State Chemistry 278, 120889 (2019).
  32. [32]B. J. Wieder, B. Bradlyn, Z. Wang, J. Cano, Y. Kim, H.-S. D. Kim, A. M. Rappe, C. L. Kane, and B. A. Bernevig, Wallpaper fermions and the nonsymmorphic Dirac insulator, Science 361, 246 (2018).DOI
  33. [33]M. V. Ale Crivillero, S. Rößler, S. Granovsky, M. Doerr, M. S. Cook, P. F. S. Rosa, J. Müller, and S. Wirth, Magnetic and electronic properties unveil polaron formation in Eu5In2Sb6, Scientific Reports 13, 1597 (2023).
  34. [34]M. V. A. Crivillero, S. Rößler, P. F. S. Rosa, J. Müller, U. K. Rössler, and S. Wirth, Surface and electronic structure at atomic length scales of the nonsymmorphic antiferromagnet Eu5In2Sb6, Physical Review B 106, 035124 (2022).
  35. [35]S. Tagliati, V. M. Krasnov, and A. Rydh, Differential membrane-based nanocalorimeter for high-resolution measurements of low-temperature specific heat, Review of Scientific Instruments 83, 055107 (2012).DOI
  36. [36]L.-L. Wang, H. C. Po, R.-J. Slager, and A. Vishwanath, Topological descendants of a multicritical Dirac semimetal with magnetism and strain, Physical Review B 104, 165107 (2021).
  37. [37]M. Pohlit, S. Rößler, Y. Ohno, H. Ohno, S. von Molnár, Z. Fisk, J. Müller, and S. Wirth, Evidence for Ferromagnetic Clusters in the Colossal-Magnetoresistance Material EuB6, Physical Review Letters 120, 257201 (2018).
  38. [38]P. Das, A. Amyan, J. Brandenburg, J. Müller, P. Xiong, S. von Molnár, and Z. Fisk, Magnetically driven electronic phase separation in the semimetallic ferromagnet EuB6, Physical Review B 86, 184425 (2012).DOI
  39. [39]S. Süllow, I. Prasad, S. Bogdanovich, M. C. Aronson, J. L. Sarrao, and Z. Fisk, Magnetotransport in the low carrier density ferromagnet EuB6, Journal of Applied Physics 87, 5591 (2000).
  40. [40]X. Zhang, L. Yu, S. von Molnár, Z. Fisk, and P. Xiong, Nonlinear Hall Effect as a Signature of Electronic Phase Separation in the Semimetallic Ferromagnet EuB6, Phys- ical Review Letters 103, 106602 (2009).
  41. [41]H. Murakawa, Y. Nakaoka, K. Iwase, T. Kida, M. Hagiwara, H. Sakai, and N. Hanasaki, Giant negative magnetoresistance in the layered semiconductor CeTe₂₋ₓSbₓ with variable magnetic polaron density, Physical Review B 107, 165138 (2023).DOI
  42. [42]R. S. Manna, P. Das, M. de Souza, F. Schnelle, M. Lang, J. Müller, S. von Molnár, and Z. Fisk, Lattice Strain Accompanying the Colossal Magnetoresistance Effect in EuB₆, Physical Review Letters 113, 067202 (2014).pubmed.ncbi.nlm.nih.gov/25148347
  43. [43]J. Chatterjee, U. Yu, and B. I. Min, Spin-polarization model: Transport properties of EuB₆, Physical Review B 69, 134423 (2004).
  44. [44]J. C. Slater, Magnetic Effects and the Hartree-Fock Equation, Physical Review 82, 538 (1951).DOI
  45. [45]Y. Tomioka, A. Asamitsu, H. Kuwahara, Y. Moritomo, M. Kasai, R. Kumai, and Y. Tokura, Magnetic-field-induced metal-insulator transition in perovskite-type manganese oxides, Physica B: Condensed Matter Proceedings of the Yamada Conference XLIV, the International Conference on the Physics of Transition Metals, 237-238, 6 (1997).DOI
  46. [46]Y. Tomioka, A. Asamitsu, Y. Moritomo, H. Kuwahara, and Y. Tokura, Collapse of a Charge-Ordered State under a Magnetic Field in Pr₁/₂Sr₁/₂MnO₃, Physical Review Letters 74, 5108 (1995).pubmed.ncbi.nlm.nih.gov/10058685
  47. [47]C. N. R. Rao, Charge, Spin, and Orbital Ordering in the Perovskite Manganates, Ln₁₋ₓAₓMnO₃ (Ln = Rare Earth, A = Ca or Sr), The Journal of Physical Chemistry B 104, 5877 (2000).DOI
  48. [48]Y.-F. Li, E. M. Been, S. Balguri, C.-J. Jia, M. B. Mahendru, Z.-C. Wang, Y. Cui, S.-D. Chen, M. Hashimoto, D.-H. Lu, B. Moritz, J. Zaanen, F. Tafti, T. P. Devereaux, and Z.-X. Shen, Colossal negative magnetoresistance from spin-polarized polarons in an ising system, Proceedings of the National Academy of Sciences 121, e2409846121.
  49. [49]K. Lai, M. Nakamura, W. Kundhikanjana, M. Kawasaki, Y. Tokura, M. A. Kelly, and Z.-X. Shen, Mesoscopic Percolating Resistance Network in a Strained Manganite Thin Film, Science 329, 190 (2010).DOI
  50. [50]M. C. Rahn, J.-R. Soh, S. Francoual, L. S. I. Veiga, J. Strempfer, J. Mardegan, D. Y. Yan, Y. F. Guo, Y. G. Shi, and A. T. Boothroyd, Coupling of magnetic order and charge transport in the candidate Dirac semimetal EuCd₂As₂, Physical Review B 97, 214422 (2018).DOI
  51. [51]S. Nasrallah, D. Santos-Cottin, F. Le Mardelé, I. Mohelský, J. Wyzula, L. Akšamović, P. Sačer, J. W. H. Barrett, W. Galloway, K. Rigaux, F. Guo, M. Puppin, I. Živković, J. H. Dil, M. Novak, C. C. Holmes, M. Orlita, N. Barišić, and A. Akrap, Magneto-optical response of the magnetic semiconductors EuCd₂X₂ (X=P,As,Sb), Physical Review B 110, L201201.
  52. [52]X. Chen, Z. Wang, Z. Zhou, W. Yang, Y. Liu, J.-Y. Lu, Z. Ren, G.-H. Cao, F. Tafti, S. Dong, and Z.-C. Wang, Manipulating magnetism and transport properties of EuCd₂P₂ with a low carrier concentration, Physical Review B 109, 224428 (2024).DOI

Paper details

vlnc:W_yvfnexzg
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents