betaFeedback

WorksPhysics and Astronomy

Thermally Hysteretic Room-Temperature Nonlinear Hall Effect in Bulk EuTe4

Abstract

The nonlinear Hall effect (NLHE) is a novel transport phenomena and shows great potential applications of future electronics. However, the realization of a giant and tunable NLHE in bulk materials at room temperature remains elusive. Here, we report a NLHE in bulk EuTe4 single crystals, which persists up to 400 K and exhibits pronounced thermal hysteresis behaviors. The ratio of nonlinear Hall voltages during warming and cooling processes is about 4.7 at 340 K. Such phenomena could be related to the inversion-symmetry-broken charge density wave state in EuTe4. These findings demonstrate bulk EuTe4_4 as a promising platform for tunable nonlinear electronic devices at room temperature.

The discovery of nonlinear Hall effect (NLHE), characterized by a transverse voltage quadratic in the applied current, has opened new avenues for probing the symmetry and topological properties of quantum materials with time-reversal symmetry, and developing novel electronic devices [1, 2, 3, 4, 5, 6, 7, 8]. Guided by the theoretical predictions [9], a Berry curvature dipole (BCD)-induced NLHE was first observed in semimetal WTe2_2 [10, 11]. However, the observed relaxation-time dependence can not be fully explained by the BCD mechanism alone, pointing to additional contributions from disorder-related processes such as skew scattering. Scaling analysis is developed to separate intrinsic and extrinsic contributions [12]. Later on, further experiments identify that quantum metric can also lead to the NLHE [13, 14].

The second-order NLHE have been observed in various systems, including topological semimetals and insulators Ta(Nb)IrTe4_4 and Bi2_2Se3_3 [10, 11, 15, 16, 17], interfacial electron gases systems like LaAlO3_3/SrTiO3_3 and CaZrO3_3/KTaO3_3 [18, 19, 20], semiconductors without inversion symmetry such as Te and CdGeAs2_2 [21, 22, 23, 24], and charge density wave (CDW) material 1T-TaS2_2 [25, 26, 27]. Moreover, the NLHE is of high potential for practical applications including radiofrequency/microwave rectification and terahertz detection [16, 21, 28, 29, 30], frequency multiplication and mixing [17, 23, 30, 31, 32], and Hall-based sensing [33, 34]. However, many of reported large NLHEs are observed mainly at low temperatures or primarily in micro- and nanoscale devices. The realization of a sizable and tunable NLHE in bulk materials at room temperature remains elusive.

The quasi-two-dimensional rare-earth telluride EuTe4_4 hosts a CDW state above the room temperature [35, 36, 37]. This CDW state breaks inversion symmetry and forms a quasi-two-dimensional (quasi-2D) polarized structure, as evidenced by room-temperature second-harmonic generation (SHG) measurements [36]. The broken inversion symmetry implies that a CDW-driven NLHE may exist in EuTe4_4. Moreover, the CDW state exhibits thermal hysteresis behaviors and can be modulated by optical and electrical stimuli [35, 38, 39, 40]. This hysteresis suggests that the NLHE might be tuned and applied for thermal memory devices [38, 39]. Here, we report a large room-temperature NLHE in bulk EuTe4_4 single crystals that persists up to 400 K. Systematic transport measurements show that the nonlinear Hall signal closely tracks the thermal hysteresis of the CDW state. Scaling analysis reveals two distinct regimes: at high temperatures, the NLHE follows the conventional scaling relation with longitudinal conductivity, whereas at lower temperatures it progressively deviates from this behavior.

EuTe4_4 possesses a quasi-2D crystal structure. It is composed of a Te-[EuTe]-Te-Te-[EuTe]-Te stacking sequence along the cc axis [Fig. 1(a)] [35], featuring both single- and double-layer Te square nets at high temperature. When entering the CDW state at lower temperature, the Te square lattices undergo structural distortion, forming Te atom trimers [Fig. 1(b)]. Crucially, the distortion and polarization of these Te trimers intrinsically break the inversion symmetry, providing the fundamental prerequisite for generating the NLHE. The intralayer polarization direction determined by the direction of structural distortion of Te trimers may be different in the single and double Te layers, which are schematically represented by the arrows in Figure 1(a). It is noted that the polarization sequence between different layers is not rigidly fixed, potentially giving rise to up to 64 distinct polarization configurations [36]. In a macroscopic sample, the stacking of different inter-layer polarization configuration lead to the formation of out-of-plane CDW domains, while different orientations of the intra-layer trimer distortions give rise to in-plane CDW domains. Despite this complexity, the macroscopic nonlinear Hall responses can be considered as a simple superposition of the contributions from multiple domains. For a domain-free Te layer, due to the constrains of crystallographic symmetry, the nonlinear transport behaviors are governed by the following formula (Supplemental Note 1),

Crystal structures, XRD pattern, and resistivity of EuTe4

Figure 1. (a) Stacking configuration of the CDW states along the cc-axis (left) and atomic positions in each Te layers along the abab-plane (right). The arrows schematically indicate the intralayer polarization directions within the single (Te 0 and Te 3) and double (Te 1/Te 2 and Te 4/Te 5) Te layers. (b) Schematic of the structural distortion in one Te layer. The left panel shows the pristine Te square lattice, and the right panel illustrates the distorted lattice in the CDW state, characterized by the formation of Te trimers. (c) XRD pattern of a EuTe4_4 single crystal. Inset: an optical image of typical EuTe4_4 single crystal. (d) Temperature dependence of zero-field ρbb(T)\rho_{bb}(T) with current applied along the bb-axis.

[Ja2ωJb2ω]=[χaaaχabb000χbba][(Eaω)2(Ebω)22EaωEbω]\begin{bmatrix} J_a^{2\omega} \\ J_b^{2\omega} \end{bmatrix} = \begin{bmatrix} \chi_{aaa} & \chi_{abb} & 0 \\ 0 & 0 & \chi_{bba} \end{bmatrix} \begin{bmatrix} (E_a^\omega)^2 \\ (E_b^\omega)^2 \\ 2E_a^\omega E_b^\omega\end{bmatrix}

where Ja2ωJ_a^{2\omega} and Jb2ωJ_b^{2\omega} are the nonlinear current densities along the aa- and bb-axes, EaωE_a^\omega and EbωE_b^\omega represent the applied alternating electric field along the aa- and bb-axes, and χaaa\chi_{aaa}, χabb\chi_{abb} and χbba\chi_{bba} are the independent non-zero elements of the nonlinear conductivity tensor. According to these symmetry constraints, an EbωE_b^\omega should produce a purely transverse nonlinear current Ja2ω=χabb(Ebω)2J_a^{2\omega}=\chi_{abb}(E_b^\omega)^2. Conversely, an EaωE_a^\omega should generate a purely longitudinal nonlinear current Ja2ω=χaaa(Eaω)2J_a^{2\omega}=\chi_{aaa}(E_a^\omega)^2. When electric field deviates from the principal crystallographic directions, there will be a finite nonlinear signal along the bb-axis. To experimentally probe these nonlinear signals, we applied an IωI^\omega along specific crystallographic axes and simultaneously measured the second-harmonic transverse and longitudinal voltages (V2ωV^{2\omega}) using lock-in amplifiers. Based on the Ohm’s law E2ω=ρ⋅J2ωE^{2\omega}=\rho\cdot J^{2\omega}, where ρ\rho is the linear resistivity tensor, the nonlinear conductivities (χabb\chi_{abb} and χaaa\chi_{aaa}) can be rigorously extracted from the measured IωI^\omega and V2ωV^{2\omega} (Supplemental Note 2).

EuTe4_4 single crystals were synthesized using the self-flux method (Supplemental Note 3). The X-ray diffraction (XRD) pattern and optical image [Figure 1(c)] confirm that the crystal surface is along the abab-plane with high quality. Figure 1(d) shows the temperature-dependent zero-field longitudinal resistivity ρbb(T)\rho_{bb}(T) along the bb-axis of EuTe4_4 single crystal. The ρbb(T)\rho_{bb}(T) exhibits a thermal hysteresis between 200 K and 400 K, consistent with previously reported results [36, 37, 41]. It indicates the presence of a CDW transition. Between 74 K – 200 K, the ρbb(T)\rho_{bb}(T) curve exhibits a metallic-like behavior without obvious hysteresis. With lowering temperature further, EuTe4_4 shows a semiconducting behavior again. This complex temperature dependence could be understood by assuming the existence of metallic CDW domain boundaries. However, these metallic pathways occupy a very small volume fraction, and thus do not pull down the bulk resistivity by orders of magnitude. Below 80 K, the metallic domain boundaries gradually become connected upon warming, and the high-resistance CDW phase dominates, leading to semiconducting behavior. Between 80 K and 360 K, the connected domain boundaries shunt the resistivity of CDW domains; as temperature rises, their metallic resistance increases, giving rise to metallic

behavior. Above 360 K, the domain boundaries network expands quickly on warming process, leading to the drop near 380 K. On cooling process, the domain boundaries contract sluggishly, causing the steep rise of resistivity near 320 K and forming the thermal hysteresis of resistivity. When cooled to 100 K – 200 K, the residual CDW regions surviving at 400 K [37, 41] act as nucleation centers to form a similar domain configuration of the warming process, restoring nearly identical longitudinal resistivity.

To probe the nonlinear transport behaviors, we measured the second-harmonic voltage V2ωV^{2\omega} as a function of the applied AC current IωI^\omega during the warming process. As depicted in Figure 2(a) and Figure 2(b), all V2ωV^{2\omega} signals display a quadratic dependence on IωI^\omega at 100 K and 300 K, which confirms the second-order nonlinear nature of the response. Crucially, the V2ωV^{2\omega} components follow the symmetry constraints at 300 K. For current applied along the bb-axis IbωI_b^\omega (sample 1) [inset of Figure 2(a)], the transverse voltage Vabb2ωV_{abb}^{2\omega} is much larger than the longitudinal one Vbbb2ωV_{bbb}^{2\omega} (∣Vabb2ω∣≫∣Vbbb2ω∣|V_{abb}^{2\omega}| \gg|V_{bbb}^{2\omega}|). Conversely, for current along the aa-axis IaωI_a^\omega (sample 2) [inset of Figure 2(b)], the longitudinal voltage Vaaa2ωV_{aaa}^{2\omega} is much larger than the transverse one Vbaa2ωV_{baa}^{2\omega} (∣Vaaa2ω∣≫∣Vbaa2ω∣|V_{aaa}^{2\omega}| \gg|V_{baa}^{2\omega}|). The residual symmetry-forbidden signals (e.g., Vbbb2ωV_{bbb}^{2\omega} and Vbaa2ωV_{baa}^{2\omega}) likely originate from slight contact misalignments or CDW domains.

Nonlinear voltage responses and temperature dependence

Figure 2. (a) Vabb2ωV_{abb}^{2\omega} and Vbbb2ωV_{bbb}^{2\omega} versus IbωI_b^\omega at 100 K and 300 K (sample 1). (b) Vbaa2ωV_{baa}^{2\omega} and Vaaa2ωV_{aaa}^{2\omega} versus IaωI_a^\omega at 100 K and 300 K (sample 2). The data for Vaaa2ωV_{aaa}^{2\omega} at 100 K are multiplied by 0.2 for clarity. Insets in (a) and (b) show the schematic figures of measuring configurations, where the red arrows indicate the direction of the applied current. (c) Vabb2ωV_{abb}^{2\omega} as a function of IbωI_b^\omega, and (d) Vaaa2ωV_{aaa}^{2\omega} as a function of IaωI_a^\omega at various temperatures. Solid lines in (a)–(d) are quadratic fits.

Furthermore, while the nonlinear signals exhibit some variation with frequency, their temperature-dependent behavior remains consistent across different frequencies [Supplemental Note 4]. This rules out artifacts such as thermal effects and parasitic capacitive coupling.

Figures 2(c) and 2(d) illustrates the temperature dependence of Vabb2ωV_{abb}^{2\omega} and Vaaa2ωV_{aaa}^{2\omega} from 100 K to 400 K during warming process. The nonlinear voltage signal robustly persists up to 400 K, and the measured V2ωV^{2\omega} can be well fitted by a quadratic curve against IωI^\omega across the entire range [Supplemental Note 5]. Notably, the Vabb2ωV_{abb}^{2\omega} [Fig. 2(c)] undergoes a distinct sign reversal between 180 K and 220 K while the Vaaa2ωV_{aaa}^{2\omega} maintains its sign across the measured temperature range [Fig. 2(d)]. This sign change suggests a temperature-driven reorientation of the polarization direction in certain layers [see Fig. 1(a)] or the evolution of intra- and interlayer domains, consistent with synchrotron X-ray diffraction experiments on interlayer CDW stacking [39, 40] and theoretical predictions of thermally driven redistribution among nearly degenerate polar CDW configurations [42]. It is noted that below T<100T < 100 K, the V2ωV^{2\omega} clearly deviates from the expected quadratic behavior against IωI^\omega. This deviation is likely attributable to capacitive effects induced by Schottky barriers or a drastic increase in contact resistance (RcR_c), which exhibits a pronounced current dependence at low temperatures [Supplemental Note 6]. Furthermore, similar to the warming process, the nonlinear transport effects during the cooling process also exhibits a significant response [Supplemental Note 7].

To correlate the nonlinear transport effects with the CDW state, we measured the V2ωV^{2\omega} as a function of IωI^\omega across a thermal cycle (3 K to 400 K to 3 K). As shown in Figure 3(a) and Figure 3(b), the (Iω)2(I^\omega)^2 dependence of V2ωV^{2\omega} exhibit distinct slopes between warming and cooling processes at 300 K. Tracking the temperature dependence [Figs. 3(c) and 3(d)] reveals a clear thermal hysteresis loop in all measured nonlinear voltage components. Throughout the entire thermal cycle, the symmetry-dictated magnitude relations (∣Vabb2ω∣≫∣Vbbb2ω∣|V_{abb}^{2\omega}| \gg|V_{bbb}^{2\omega}| and ∣Vaaa2ω∣≫∣Vbaa2ω∣|V_{aaa}^{2\omega}| \gg|V_{baa}^{2\omega}|) are preserved in principle. Such hysteresis behaviors of V2ω(T)V^{2\omega}(T) undoubtedly indicate that the nonlinear transport phenomena are intimately related to the CDW state, which leads to the hysteresis of ρbb(T)\rho_{bb}(T) [Fig. 1(d)]. Moreover, the convergence temperature of the hysteresis is close to that of the linear resistivity, further confirming that the observed nonlinear voltage signals are intrinsic. The ratios Vabb2ωV_{abb}^{2\omega}(warming)/Vabb2ωV_{abb}^{2\omega}(cooling) and Vaaa2ωV_{aaa}^{2\omega}(warming)/Vaaa2ωV_{aaa}^{2\omega}(cooling) reach their maximum values of approximately 4.7 and 4.9 at 340 K, respectively [Figs. 3(c) and 3(d)].

Thermal hysteresis of the nonlinear transport response

Figure 3. Thermal hysteresis of the nonlinear transport response. (a) Vabb2ωV_{abb}^{2\omega} and Vbbb2ωV_{bbb}^{2\omega}, and (b) Vaaa2ωV_{aaa}^{2\omega} and Vbaa2ωV_{baa}^{2\omega} versus (Iω)2(I^\omega)^2 measured at 300 K during warming and cooling processes. In (a) and (b), the (Iω)2(I^\omega)^2 spans from 0 mA2^2 to 25 mA2^2. Temperature dependence of (c) Vabb2ωV_{abb}^{2\omega} and Vbbb2ωV_{bbb}^{2\omega}, and (d) Vbaa2ωV_{baa}^{2\omega} and Vaaa2ωV_{aaa}^{2\omega} during the thermal cycle. Arrows in (c) and (d) indicate the temperature sweep directions.

The temperature-dependent χabb\chi_{abb} and χaaa\chi_{aaa} are presented in Figure 4(a) and Figure 4(b) and Supplemental Note 8. Both χabb\chi_{abb} and χaaa\chi_{aaa} exhibit pronounced thermal hysteresis behaviors above about 250 K, similar to those hysteresis of V2ω(T)V^{2\omega}(T) and ρbb(T)\rho_{bb}(T). To elucidate the microscopic origin of the NLHE in EuTe4_4, the scaling law analysis is performed. The nonlinear resistivity is generally expressed as [11, 12]

Temperature dependence, scaling-law analysis, and comparison of nonlinear responsivity

Figure 4. (a) Temperature dependence of χabb(T)\chi_{abb}(T) measured during the warming and cooling processes between 100 K and 400 K. (b) Scaling law analysis of Eabb2ω/Eb2E^{2\omega}_{abb}/E_b^2 versus σbb2\sigma_{bb}^2. The inset displays an enlarged view of the data near the origin. (c) Comparison of the nonlinear responsivity ∣E⊥ω∣/(Eω)2\lvert E^\omega_{\perp}\rvert/(E^\omega)^2 as a function of temperature between bulk EuTe4_4 and most of reported materials exhibiting the NLHE [10, 11, 13, 14, 15, 16, 17, 18, 21, 23, 24, 26, 30, 32, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63]. Red circles and blue squares denote EuTe4_4 during warming and cooling, respectively. Literature data are classified into topological materials, designed platforms, and other materials (see legend). The light-blue shaded region highlights the high-temperature and high-responsivity region. Details and conditions are given in Supplemental Note 10.

Eabb2ωEb2=ξσbb2+η\frac{E^{2\omega}_{abb}}{E_b^2} = \xi\sigma_{bb}^2 + \eta

The ξσbb2\xi\sigma_{bb}^2 term is attributed to skew scattering, while the η\eta term originates from the intrinsic contribution (Berry curvature dipole) and side-jump scattering. As shown in Figure 4(b), in the high-temperature regime (280 K–400 K for the warming process and 240 K–380 K for the cooling process), the Eabb2ω/Eb2E^{2\omega}_{abb}/E_b^2 and σbb2\sigma_{bb}^2 exhibits a clear linear relationship, validating the scaling law. The linear fits yield the intercepts η\eta of −2.7(1)×10−4-2.7(1)\times10^{-4} and −3.2(3)×10−4-3.2(3)\times10^{-4} V−1^{-1} cm, and the slopes ξ\xi of −8.2(2)×10−8-8.2(2)\times10^{-8} and −5.1(3)×10−8-5.1(3)\times10^{-8} V−1^{-1} cm3^3 Ω\Omega for the warming and cooling processes, respectively (Supplemental Note 9). It indicates that the extrinsic mechanism (skew scattering) should dominate the nonlinear transport behaviors at high-temperature region. Such a negligibly small intrinsic contribution is physically reasonable. Previous transport and ARPES studies have established that EuTe4_4 exhibits a fully gapped semiconductor behavior, with a maximum energy gap EgE_g of approximately 196 meV at 55 K [37]. Since the intrinsic contribution (Berry curvature dipole) is governed by the integral of the Berry curvature derivatives over the states at the Fermi energy level, the absence of a Fermi surface in this CDW-gapped state naturally dictates that the Berry curvature dipole should be vanishingly small. The finite contribution of the intrinsic mechanism could arise from thermally excited carriers at high temperature, as well as in-gap states induced by defects or domain boundaries in the extrinsic semiconductor regime, which typically inherit the character of their parent bands. Furthermore, the linear fits reveal a noticeable difference in the slope ξ\xi between warming and cooling processes. Since ξ\xi represents the skew scattering coefficient, it indicates that scattering process may change due to the different configurations of CDW domains between warming and cooling processes.

In contrast to the high-temperature regime, the low-temperature data deviate from the scaling law, particularly during the cooling process between 120 K and 220 K and the warming process between 180 K and 260 K [inset of Figure 4(b)]. Furthermore, this deviation is accompanied by a sign change of Eabb2ω/Eb2E^{2\omega}_{abb}/E_b^2. There are one possible explanation of the observed breakdown of scaling behavior and sign reversal behavior. In the high-temperature regime (>200> 200 K), the macroscopic domain evolution is primarily characterized by the transition between CDW and metallic states as well as the merging of domains with different polarization directions. This process concurrently modulates both of mobility and nonlinear Hall response, thus preserving the validity of the scaling behavior. However, at low temperatures, the domain evolution mainly involves the realignment of polarization directions across different layers. While such 180∘180^\circ polarization flips leave the in-plane mobility largely unaffected, they profoundly alter the symmetry-dependent nonlinear Hall signal, naturally leading to a breakdown of the scaling behavior and the corresponding sign reversal. Figure 4(c) and Supplemental Note 10 show the nonlinear responsivity ∣E⊥2ω∣/(Eω)2\lvert E^{2\omega}_{\perp}\rvert/(E^\omega)^2 as a function of temperature for materials exhibiting the NLHE. In most of reported materials, such as topological semimetals, large nonlinear responsivities only appear at low temperatures (below 30 K) and drop quickly as the temperature increases toward room temperature. In contrast, bulk EuTe4_4 is located in the high-temperature and high-responsivity region (T≥100T \geq100 K and ∣E⊥2ω∣/(Eω)2≥10−1\lvert E^{2\omega}_{\perp}\rvert/(E^\omega)^2 \geq10^{-1} μ\muV V−1^{-1}). In contrast, bulk EuTe4_4 maintains a giant nonlinear responsivity across a wide temperature range up to 400 K. At room temperature, its nonlinear responsivity ∣E⊥2ω∣/(Eω)2\lvert E^{2\omega}_{\perp}\rvert/(E^\omega)^2 reaches ∼3\sim3–6 μ\muV V−1^{-1} (and up to ∼15\sim15 μ\muV V−1^{-1} at 400 K), which is several orders of magnitude higher than most materials at room temperature, such as TaIrTe4_4 [16], few-layer WTe2_2 [10, 11], and strained monolayer WSe2_2 [46]. More importantly, unlike other materials with obvious room-temperature NLHE, the nonlinear responsivity in EuTe4_4 exhibits a pronounced thermal hysteresis loop between warming and cooling processes, providing an extra tunability of NLHE.

EuTe4_4 provides a promising platform to explore the novel properties and tunability of CDW-induced nonlinear electronic transport effects. First, fabricating micro- or nano-scale devices could isolate single CDW domain, thereby circumventing the averaging effect and yielding even higher nonlinear signals that directly reflect the intrinsic CDW state. Second, the CDW domains and their polarization directions could be tuned by strain, and the carrier density and the associated CDW phases in semiconducting EuTe4_4 thin flakes can also be effectively tuned via electrostatic gating. Thus nonlinear transport effects may also be manipulated by these external fields effectively. Combining this tunable giant nonlinear response with the pronounced thermal hysteresis origin to the CDW state, EuTe4_4 demonstrates tremendous potential for advanced applications of NLHE, particularly in novel memory devices.

In summary, we have systematically investigated the nonlinear transport effects of bulk EuTe4_4 crystals. Our results reveal that the nonlinear Hall voltage is strongly coupled to the hysteretic CDW state, leading to a pronounced thermal memory effect that can modulate the signal by a factor of about 4.7 at 340 K. While the high-temperature transport conforms to standard scaling law, the continuous dynamic reconfiguration of CDW domains at lower temperatures may lead to a breakdown of scaling behavior. Despite this complex domain configuration, EuTe4_4 sustains a large nonlinear Hall response up to 400 K, with room-temperature responsivities exceeding representative reported values for classic TaIrTe4_4 and few-layer WTe2_2 systems. Our findings demonstrate that EuTe4_4 is a highly promising platform for designing novel, tunable nonlinear electronic devices at room temperature.

This work was supported by the National Key R&D Program of China (Grant Nos. 2023YFA1406500 and 2022YFA1403800), the National Natural Science Foundation of China (Grant Nos. 12625409, 12274459 and 52572288, and W2431011), the Shanghai Key Laboratory of Novel Extreme Condition Materials, China (Grant No. 22dz2260800), the Shanghai Science and Technology Committee, China (Grant No. 22JC1410300), the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China (Grant No. 26XNKJ22).

Supplementary Information for Thermally Hysteretic Room-Temperature Nonlinear Hall Effect in Bulk EuTe4_4

Yang Fu; Xingyu Lv; Shangjie Tian; Ying Ma; Xiao Zhang; Cedomir Petrovic; Hechang Lei

Symmetry Analysis and Tensor Symmetries

The room-temperature charge-density-wave (CDW) phase of bulk EuTe4_4 belongs to the non-centrosymmetric space group P21cnP2_1cn (No. 33) [1]. For macroscopic second-order nonlinear transport responses, the relevant symmetry is determined by the crystallographic point group. In this context, microscopic translation-related symmetry operations, such as glide planes and screw axes, act effectively as simple mirror planes and rotation axes. Consequently, the nonlinear transport properties of EuTe4_4 are governed by the point group C2vC_{2v} (mm2mm2), which is identical to the point group of the well-known nonlinear Hall material WTe2_2 and TaIrTe4_4 [2, 3].

To derive the symmetry-allowed nonlinear conductivity tensor, we begin with the general phenomenological equation for the second-order nonlinear current response,

Ji2ω=∑j,kχijkEjωEkω(1)J_i^{2\omega}=\sum_{j,k}\chi_{ijk}E_j^\omega E_k^\omega \tag*{(1)}

where the indices i,j,k∈{a,b,c}i,j,k \in\{a,b,c\} denote the crystallographic directions. Here, Ji2ωJ_i^{2\omega} is the ii-th component of the second-order nonlinear current density, EjωE_j^\omega and EkωE_k^\omega are the components of the applied alternating electric field at the frequency ω\omega, and χijk\chi_{ijk} represents the elements of the nonlinear conductivity tensor. Because both the current density J2ωJ^{2\omega} and the electric field EωE^\omega are polar vectors, the nonlinear conductivity χijk\chi_{ijk} transforms as a third-rank polar tensor. Furthermore, since the electric fields commute (EjωEkω=EkωEjωE_j^\omega E_k^\omega=E_k^\omega E_j^\omega), the tensor χijk\chi_{ijk} is symmetric with respect to its last two indices (χijk=χikj\chi_{ijk}=\chi_{ikj}). This allows us to express the tensor as a 3×63 \times6 matrix with 18 independent elements (Voigt notation),

[Ja2ωJb2ωJc2ω]=[χaaaχabbχaccχabcχaacχaabχbaaχbbbχbccχbbcχbacχbabχcaaχcbbχcccχcbcχcacχcab][(Eaω)2(Ebω)2(Ecω)22EbωEcω2EaωEcω2EaωEbω](2)\begin{bmatrix} J_a^{2\omega}\\ J_b^{2\omega}\\ J_c^{2\omega} \end{bmatrix} = \begin{bmatrix} \chi_{aaa} & \chi_{abb} & \chi_{acc} & \chi_{abc} & \chi_{aac} & \chi_{aab}\\ \chi_{baa} & \chi_{bbb} & \chi_{bcc} & \chi_{bbc} & \chi_{bac} & \chi_{bab}\\ \chi_{caa} & \chi_{cbb} & \chi_{ccc} & \chi_{cbc} & \chi_{cac} & \chi_{cab} \end{bmatrix} \begin{bmatrix} (E_a^\omega)^2\\ (E_b^\omega)^2\\ (E_c^\omega)^2\\ 2E_b^\omega E_c^\omega\\ 2E_a^\omega E_c^\omega\\ 2E_a^\omega E_b^\omega \tag*{(2)} \end{bmatrix}

Based on the crystal symmetry, the space group P21cnP2_1cn possesses the following key symmetry operations:

  1. A glide plane perpendicular to the bb-axis, which effectively acts as a mirror plane (MbM_b), transforming the coordinates as (a,b,c)→(a,−b,c)(a,b,c) \to(a,-b,c).

  1. A glide plane perpendicular to the cc-axis, which effectively acts as a mirror plane (McM_c), transforming the coordinates as (a,b,c)→(a,b,−c)(a,b,c) \to(a,b,-c).

  1. A two-fold screw axis along the aa-axis, which effectively acts as a rotation axis (C2aC_{2a}), transforming the coordinates as (a,b,c)→(a,−b,−c)(a,b,c) \to(a,-b,-c).

The nonlinear conductivity tensor elements must remain invariant under these symmetry operations. Because the operations involve only reflections or two-fold rotations along the principal axes, each spatial index is either preserved or inverted. Consequently, a transformed tensor element χi′j′k′\chi_{i'j'k'} is equal to ±χijk\pm\chi_{ijk}, where the negative sign applies if the operation inverts an odd number of its corresponding spatial coordinates. To satisfy the invariance requirement (χi′j′k′=χijk\chi_{i'j'k'}=\chi_{ijk}), any element that acquires a negative sign must be identically zero. For example, under the MbM_b mirror symmetry, any tensor element with an odd number of bb indices (such as χaab\chi_{aab}) transforms to its negative (χi′j′k′=−χaab\chi_{i'j'k'}=-\chi_{aab}), and therefore must be zero. Applying these parity rules to all 18 elements, we find that the only non-zero components are χaaa\chi_{aaa}, χabb\chi_{abb}, χacc\chi_{acc}, χab=χbba\chi_{ab}=\chi_{bba}, and χcac=χcca\chi_{cac}=\chi_{cca}. Thus, the macroscopic nonlinear transport equation can be elegantly written in the following matrix form for space group P21cnP2_1cn,

[Ja2ωJb2ωJc2ω]=[χaaaχabbχacc00000000χbba0000χcca0][(Eaω)2(Ebω)2(Ecω)22EbωEcω2EaωEcω2EaωEbω](3)\begin{bmatrix} J_a^{2\omega}\\ J_b^{2\omega}\\ J_c^{2\omega} \end{bmatrix} = \begin{bmatrix} \chi_{aaa} & \chi_{abb} & \chi_{acc} & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0 & \chi_{bba}\\ 0 & 0 & 0 & 0 & \chi_{cca} & 0 \end{bmatrix} \begin{bmatrix} (E_a^\omega)^2\\ (E_b^\omega)^2\\ (E_c^\omega)^2\\ 2E_b^\omega E_c^\omega\\ 2E_a^\omega E_c^\omega\\ 2E_a^\omega E_b^\omega \tag*{(3)} \end{bmatrix}

When the driving electric field is confined within the abab-plane (Ecω=0E_c^\omega=0), the formula further can be simplified as following,

[Ja2ωJb2ωJc2ω]=[χaaaχabb000000000χbba000000][(Eaω)2(Ebω)20002EaωEbω](4)\begin{bmatrix} J_a^{2\omega}\\ J_b^{2\omega}\\ J_c^{2\omega} \end{bmatrix} = \begin{bmatrix} \chi_{aaa} & \chi_{abb} & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0 & \chi_{bba}\\ 0 & 0 & 0 & 0 & 0 & 0 \end{bmatrix} \begin{bmatrix} (E_a^\omega)^2\\ (E_b^\omega)^2\\ 0\\ 0\\ 0\\ 2E_a^\omega E_b^\omega \tag*{(4)} \end{bmatrix}

For a single Te layer, the relevant in-plane symmetry reduces to a single mirror plane. Under this mirror symmetry constraint, the two-dimensional nonlinear transport matrix is given by

[Ja2ωJb2ω]=[χaaaχabb000χbba][(Eaω)2(Ebω)22EaωEbω](5)\begin{bmatrix} J_a^{2\omega}\\ J_b^{2\omega} \end{bmatrix} = \begin{bmatrix} \chi_{aaa} & \chi_{abb} & 0\\ 0 & 0 & \chi_{bba} \end{bmatrix} \begin{bmatrix} (E_a^\omega)^2\\ (E_b^\omega)^2\\ 2E_a^\omega E_b^\omega \tag*{(5)} \end{bmatrix}

The macroscopic nonlinear effect is ultimately the superposition of contributions from all individual Te layers. To generalize the response for an arbitrary in-plane measurement geometry, suppose the driving electric field EωE^\omega is applied within the abab-plane at an angle θ\theta relative to the aa-axis. The electric field components are Eaω=Eωcos⁡θE_a^\omega=E^\omega\cos\theta and Ebω=Eωsin⁡θE_b^\omega=E^\omega\sin\theta. Substituting these into the simplified matrix yields the nonlinear current components along the principal axes,

Ja2ω=(Eω)2(χaaacos⁡2θ+χabbsin⁡2θ)J_a^{2\omega}=(E^\omega)^2(\chi_{aaa}\cos^2\theta+\chi_{abb}\sin^2\theta)
{#eq:S6}

J_b^{2\omega}=2(E^\omega)^2\chi_{bba}\sin\theta\cos\theta

Equation not extracted

The corresponding longitudinal (J∥2ωJ_{\parallel}^{2\omega}) and transverse (J⊥2ωJ_{\perp}^{2\omega}) nonlinear currents measured relative to the applied field direction can be obtained through coordinate rotation,

J∥2ω=Ja2ωcos⁡θ+Jb2ωsin⁡θ=(Eω)2cos⁡θ[χaaacos⁡2θ+(χabb+2χbba)sin⁡2θ]J⊥2ω=−Ja2ωsin⁡θ+Jb2ωcos⁡θ=(Eω)2sin⁡θ[(2χbba−χaaa)cos⁡2θ−χabbsin⁡2θ](6)\begin{aligned} J_{\parallel}^{2\omega}&=J_a^{2\omega}\cos\theta+J_b^{2\omega}\sin\theta=(E^\omega)^2\cos\theta\left[\chi_{aaa}\cos^2\theta+(\chi_{abb}+2\chi_{bba})\sin^2\theta\right]\\ J_{\perp}^{2\omega}&=-J_a^{2\omega}\sin\theta+J_b^{2\omega}\cos\theta=(E^\omega)^2\sin\theta\left[(2\chi_{bba}-\chi_{aaa})\cos^2\theta-\chi_{abb}\sin^2\theta\right] \tag*{(6)} \end{aligned}

where J∥2ωJ_{\parallel}^{2\omega} and J⊥2ωJ_{\perp}^{2\omega} are the nonlinear current densities parallel and perpendicular to the applied electric field EωE^\omega, respectively, θ\theta represents the angle between the electric field and the crystallographic aa-axis, and χaaa\chi_{aaa}, χabb\chi_{abb}, and χbba\chi_{bba} are the independent non-zero tensor elements of the nonlinear conductivity dictated by the P21cnP2_1cn symmetry.

Crucially, these equations demonstrate that any slight misalignment of the electrical contacts relative to the principal crystallographic axes will induce residual symmetry-forbidden signals. For instance, if the applied electric field is nominally directed along the aa-axis but suffers from a small misalignment angle θ\theta, the transverse nonlinear current J⊥2ωJ_{\perp}^{2\omega} (which corresponds to the symmetry-forbidden Vbaa2ωV_{baa}^{2\omega} signal in the main text) is no longer zero, but becomes linearly proportional to the misalignment,

J⊥2ω(θ→0)≈(Eω)2θ(2χbba−χaaa)(7)J_{\perp}^{2\omega}(\theta\to0)\approx(E^\omega)^2\theta(2\chi_{bba}-\chi_{aaa}) \tag*{(7)}

Similarly, if the electric field is nominally applied along the bb-axis (corresponding to θ≈π/2\theta\approx\pi/2), we can define a small deviation angle δ=π/2−θ\delta=\pi/2-\theta. The longitudinal nonlinear current J∥2ωJ_{\parallel}^{2\omega} (which corresponds to the symmetry-forbidden Vbbb2ωV_{bbb}^{2\omega} signal and is exactly zero when perfectly aligned) emerges as

J∥2ω(δ→0)≈(Eω)2δ(χabb+2χbba)(8)J_{\parallel}^{2\omega}(\delta\to0)\approx(E^\omega)^2\delta(\chi_{abb}+2\chi_{bba}) \tag*{(8)}

Therefore, the inevitable small deviations in device fabrication or crystal cleavage naturally account for the finite residual symmetry-forbidden nonlinear voltages observed in the transport measurements.

Calculation of Nonlinear Conductivities

In transport measurements, the second-harmonic signals are detected under near open-circuit conditions. Under these conditions, the intrinsic nonlinear current density J2ω\mathbf{J}^{2\omega} generates a corresponding nonlinear electric field E2ω\mathbf{E}^{2\omega} to maintain zero net current, which satisfies the Ohm’s law,

E2ω=ρ⋅J2ω(9)\mathbf{E}^{2\omega} = \boldsymbol{\rho} \cdot\mathbf{J}^{2\omega} \tag*{(9)}

where ρ\boldsymbol{\rho} is the linear resistivity tensor for anisotropic EuTe4_4,

ρ=(ρa000ρb000ρc)(10)\boldsymbol{\rho} = \begin{pmatrix} \rho_a & 0 & 0 \\ 0 & \rho_b & 0 \\ 0 & 0 & \rho_c \end{pmatrix} \tag*{(10)}

Thus, the components of the intrinsic nonlinear current density can be deduced as Ji2ω=Ei2ω/ρiJ_i^{2\omega} = E_i^{2\omega}/\rho_i (for i=a,b,ci = a, b, c). Both the applied primary electric field and the generated nonlinear electric field intensities are obtained by dividing their corresponding measured voltages by the distance between the respective measurement electrodes.

To experimentally determine the nonlinear conductivities, we express the applied primary electric field EωE^\omega using the applied alternating current IωI^\omega, the distance between the longitudinal voltage terminals ll, the sample width ww, and the sample thickness dd. When the primary current is applied along the bb-axis, the electric field is Ebω=ρbjbω=Iωρb/(wd)E_b^\omega= \rho_b j_b^\omega= I^\omega\rho_b/(wd). Using the transverse nonlinear current Ja2ω=Ea2ω/ρaJ_a^{2\omega} = E_a^{2\omega}/\rho_a and the transverse nonlinear electric field Ea2ω=Vabb2ω/wE_a^{2\omega} = V_{abb}^{2\omega}/w, and assuming ρa≈ρb\rho_a \approx\rho_b, the transverse (nonlinear Hall) conductivity χabb=Ja2ω/(Ebω)2\chi_{abb} = J_a^{2\omega}/(E_b^\omega)^2 can be expressed as

χabb=Vabb2ωwd2(Ibω)2ρbρa≈Vabb2ωwd2(Ibω)2ρb2(11)\chi_{abb} = \frac{V_{abb}^{2\omega}wd^2}{(I_b^\omega)^2\rho_b\rho_a} \approx\frac{V_{abb}^{2\omega}wd^2}{(I_b^\omega)^2\rho_b^2} \tag*{(11)}

Similarly, when the primary current is applied along the aa-axis, the electric field is Eaω=ρajaω=Iaωρa/(wd)E_a^\omega= \rho_a j_a^\omega= I_a^\omega\rho_a/(wd). Using the longitudinal nonlinear current Ja2ω=Ea2ω/ρaJ_a^{2\omega} = E_a^{2\omega}/\rho_a and Ea2ω=Vaaa2ω/lE_a^{2\omega} = V_{aaa}^{2\omega}/l, the longitudinal nonlinear conductivity χaaa=Ja2ω/(Eaω)2\chi_{aaa} = J_a^{2\omega}/(E_a^\omega)^2 can be expressed as

χaaa=Vaaa2ωwd2(Iaω)2ρa3l(12)\chi_{aaa} = \frac{V_{aaa}^{2\omega}wd^2}{(I_a^\omega)^2\rho_a^3 l} \tag*{(12)}

Crystal Growth and Experimental Methods

EuTe4_4 single crystals are grown using the self-flux method. Eu chunks (99.9 %) and Te chunks (99.999 %) were mixed in a molar ratio of 1 : 15 and loaded into a quartz crucible, which was then sealed in a quartz ampoule under high vacuum (10−210^{-2} Pa). The sealed ampoule was heated to 1123 K over 24h, maintained there for 48h, then cooled at a rate of 4 K/h to 693 K. After maintaining this temperature for one week, the mixture was centrifuged to remove flux and the plate-like EuTe4_4 single crystals can be obtained.

The devices measured along the two different current directions along the crystallographic axes (as discussed in the main text) were cut from the same one single crystal. The crystallographic axes of the crystals were accurately determined using a X-ray diffractometer equipped with a goniometer (Bruker D8 Advance). After carefully cleaning the crystal surfaces, platinum (Pt) electrical contacts were deposited via magnetron sputtering through a hard mask. Pt wires were then attached using conductive epoxy and baked inside a glove box to form robust and clean electrical connections.

For the transport measurements, an AC current was applied using a Keithley 6221 current source, and the corresponding primary and harmonic voltages were recorded using two NF LI5650 lock-in amplifiers. When measuring the contact resistance (RcR_c), one LI5650 lock-in amplifier was utilized as the AC voltage source. The applied current was simultaneously monitored by measuring the voltage drop across a 1 kΩ\Omega series resistor using the other lock-in amplifier. The RcR_c was subsequently derived from the ratio of the output voltage of the LI5650 to the measured current. All temperature-dependent transport measurements were conducted in a home-built high-vacuum cryogenic system, enabling precise temperature control over a wide range from 3 K up to 400 K.

Frequency Dependence of the Nonlinear Signal

To ensure that the observed nonlinear signals originate from the intrinsic physical properties of EuTe4\mathrm{EuTe}_{4} rather than parasitic artifacts, we investigated the frequency dependence of the second-harmonic responses. As shown in Figure S1, the Vabb2ωV^{2\omega}_{abb} exhibits a slight frequency dependence, but its overall behavior remains highly consistent. This indicates that the intrinsic nonlinear signal dominates the voltage response, thereby largely ruling out extrinsic contributions such as parasitic capacitive coupling and local Joule heating. In contrast, the Vaaa2ωV^{2\omega}_{aaa} displays a relatively stronger frequency dependence compared to the Vabb2ωV^{2\omega}_{abb}. This difference arises because the longitudinal measurement is more susceptible to parasitic capacitance and thermal gradient effects. Specifically, Joule heating (∝(Iω)2\propto(I^\omega)^2) primarily generates longitudinal thermal gradients, which produce a significant 2ω2\omega thermoelectric voltage along the primary current path. Consequently, the Vabb2ωV^{2\omega}_{abb} is largely unaffected by these longitudinal thermal gradients and serves as a more reliable probe for the nonlinear Hall effect. Nevertheless, despite this moderate frequency dependence in the longitudinal channel, the overall temperature-dependent behavior remains consistent across the measured frequency range.

Frequency dependence of the second-harmonic voltages

Figure S1. Frequency dependence of the second-harmonic voltages. Temperature-dependent (a) Vabb2ωV^{2\omega}_{abb}, (b) Vbbb2ωV^{2\omega}_{bbb}, (c) Vaaa2ωV^{2\omega}_{aaa}, and (d) Vbaa2ωV^{2\omega}_{baa} under various frequencies. The black arrows indicate the temperature sweep directions. The inset in (c) shows the enlarged view of Vaaa2ωV^{2\omega}_{aaa} at the high-temperature region.

High-Temperature Nonlinear Hall Effect

As discussed in the main text, the nonlinear transport response in EuTe4_4 exhibits a robust quadratic dependence on the applied current IωI^{\omega} across the high-temperature range. Figure S2 displays the detailed Vabb2ωV_{abb}^{2\omega} curves measured during the warming process from 300 K to 400 K, confirming that the nonlinear signal is well fitted by a quadratic relationship within this high-temperature regime.

Nonlinear Hall effect between 300 K–400 K

Figure S2. Nonlinear Hall effect between 300 K–400 K. (a) Vabb2ωV_{abb}^{2\omega} as a function of IbωI_b^{\omega} during the warming process from 300 K to 400 K. The solid lines represent the quadratic fits.

Supplemental Note 6: Nonlinear Hall Effect and Contact Resistance at Low Temperatures

As discussed in the main text, while the second-harmonic voltage strictly adheres to the expected quadratic dependence on the applied current (IbωI_b^\omega) at high temperatures, this relationship begins to break down upon cooling below 100 K. As shown in Figure S3 and Figure S3, the nonlinear Hall voltages measured during the warming process exhibit a clear deviation from the typical quadratic behavior at low temperatures. To elucidate the origin of this deviation, we examined the current dependence of RcR_\mathrm{c} at various temperatures, as shown in Figure S3 and Figure S3. Upon cooling below 40 K, the RcR_\mathrm{c} increases drastically, reaching several kilo-ohms. More importantly, it exhibits a pronounced current dependence, decreasing sharply as the current increases. This strong current dependence is a characteristic signature of the formation of Schottky barriers at the interface between the metallic electrodes and the EuTe4_4 crystal. In contrast, at higher temperatures (≥100\ge100 K) [Figure S3], the RcR_\mathrm{c} is significantly lower (around 150 Ω\Omega) and becomes basically independent of the applied current, confirming the presence of good Ohmic contacts. The emergence of Schottky barriers at low temperatures could induce parasitic capacitive effects and non-ideal transport behaviors, which contaminate the intrinsic bulk nonlinear Hall response and cause the deviation from the quadratic scaling law. Therefore, to ensure the reliability of our analysis, the scaling law discussions in the main text are strictly confined to the high-temperature regime (≥100\ge100 K) with Ohmic contacts.

Four-panel plots of nonlinear Hall voltage and contact resistance at low temperatures

Figure S3. Nonlinear Hall effect and contact resistance at low temperatures. Second-harmonic voltage Vabb2ωV_{abb}^{2\omega} versus the applied current IbωI_b^\omega measured during the warming process at (a) 3 K to 80 K and (b) 60 K to 140 K. The solid lines in (b) are quadratic fits, showing a clear deviation from the expected quadratic behavior below 100 K. Current dependence of RcR_\mathrm{c} at (c) 3 K to 80 K and (d) 60 K to 140 K. At low temperatures, RcR_\mathrm{c} increases drastically and exhibits a pronounced nonlinear dependence on the applied current, indicative of the formation of Schottky barriers, whereas it is small and essentially current-independent at higher temperatures.

Nonlinear Hall Effect During the Cooling Process

To further investigate the nonlinear transport behaviors associated with the CDW transition, we also measured the second-harmonic voltages during the cooling process (from 400 K down to 3 K). As shown in Figure S4, the nonlinear Hall effect during the cooling process exhibits a similar quadratic current dependence at high temperatures. Furthermore, upon cooling below 100 K, the nonlinear response again deviates from the typical quadratic scaling, consistent with the emergence of Schottky barriers and the increase of contact resistance at low temperatures.

Four-panel plots of second-harmonic voltages during the cooling process

Figure S4. Nonlinear Hall effect during the cooling process. Second-harmonic voltages (a) Vabb2ωV_{abb}^{2\omega} and (b) Vaaa2ωV_{aaa}^{2\omega} as a function of the applied current IωI^{\omega} from 100 K to 380 K. The solid lines in (a) and (b) represent quadratic fits. (c) Vabb2ωV_{abb}^{2\omega} and (d) Vaaa2ωV_{aaa}^{2\omega} as a function of IωI^{\omega} from 3 K to 80 K. In (d), the data for Vaaa2ωV_{aaa}^{2\omega} at 3 K are multiplied by 0.3 for clarity. The nonlinear signals exhibit a typical quadratic behavior at high temperatures and deviate at low temperatures.

Temperature Dependence of Longitudinal Nonlinear Conductivity χaaa\chi_{aaa}

Temperature dependence of $\chi_{aaa}(T)$ during warming and cooling

Figure S5. Temperature dependence of χaaa(T)\chi_{aaa}(T) measured during the warming and cooling processes between 100 K and 400 K.

SUPPLEMENTAL NOTE 9: DETAILED FITTING CURVES FOR SCALING LAW

Linear fits of the scaling law

Figure S6. Linear fits of the scaling law Eabb/Eb2E_{abb}/E_b^2 versus σbb2\sigma_{bb}^2 at high temperatures. The red circles and black squares represent the experimental data for the warming (280 K–400 K) and cooling (240 K–380 K) branches, respectively. The solid lines denote the corresponding linear fits.

Comparison of Nonlinear Transport Parameters

MaterialTT (K)E⊥2ω/(E∥ω)2E_{\perp}^{2\omega}/(E_{\parallel}^{\omega})^2 (μm V−1\mu\mathrm{m}\,\mathrm{V}^{-1})RemarksRef.
EuTe4_440014.76This work
EuTe4_43003.581Warming branchThis work
EuTe4_43006.108Cooling branchThis work
EuTe4_410054.1Warming branchThis work
EuTe4_410011.15Cooling branchThis work
WTe2_2 {sup}3`10152 layers[2]
WTe2_2 {sup}2`100.001415 layers[4]
BP/MnBi2_2Te4_4/BP {sup}3`1.82.36 SL; Gated[5]
BP/MnBi2_2Te4_4/BP {sup}3`110.86 SL[5]
MnBi2_2Te4_4 {sup}2`22.74 SL[6]
MnBi2_2Te4_4 {sup}2`24.86 SL[6]
TaIrTe4_4 {sup}2`3000.0047t=16t = 16 nm[3]
TaIrTe4_4 {sup}2`20.0248t=16t = 16 nm[3]
Bi2_2Se3_3 {sup}2`100.0011520 QL[7]
Bi2_2Se3_3 {sup}2`2003.9×10−43.9 \times10^{-4}20 QL[7]
TdT_d-MoTe2_2 {sup}1`24.3×1054.3 \times10^5t=127t = 127 nm; cc-axis output[8]
Ce3_3Bi4_4Pd3_3 {sup}1`≈1.7\approx1.73×1033 \times10^3[9]
WSe2_2 {sup}3`1.51×1041 \times10^4Twisted bilayer; Gated[10]
Graphene {sup}2`1.75.1×1035.1 \times10^3Twisted bilayer; Gated[11]
Graphene {sup}2`804.6Twisted bilayer; Gated[11]
WSe2_2 {sup}2`1400.028Strained monolayer; Gated[12]
WTe2_2 {sup}2`59112 layers; in-plane d.c. electric field[13]
BaMnSb2_2 {sup}2`300170[14]
BaMnSb2_2 {sup}2`2061.3×1031.3 \times10^3[14]
LaAlO3_3/SrTiO3_3 {sup}2`1.52.6(111) interface; Gated[15]
Bi {sup}2`3000.013t=100t = 100 nm[16]
Bi {sup}2`3330.019t=100t = 100 nm[16]
Te {sup}2`3000.075Gated[17]
Te {sup}2`3000.015[17]
Te {sup}2`2000.075[17]
PbTaSe2_2 {sup}2`3000.0118t=123t = 123 nm[18]
PbTaSe2_2 {sup}2`2.81.45×1041.45 \times10^4t=123t = 123 nm; superconducting mixed state[18]
PbTaSe2_2 {sup}2`6.570t=123t = 123 nm; normal state[18]
ZrTe5_5 {sup}3`20.75t=60t = 60 nm[19]
BiTeBr {sup}1`3000.57t=4t = 4 nm[20]
BiTeBr {sup}2`207.2t=4t = 4 nm[20]
BiTeBr {sup}2`3500.42t=4t = 4 nm[20]

Table S1. Nonlinear Hall responsivity of EuTe4_4 and typical materials reported in the literature.

MaterialTT (K)E⊥2ω/(E∥ω)2E_{\perp}^{2\omega}/(E_{\parallel}^{\omega})^{2} (μm V−1\mu\mathrm{m}\,\mathrm{V}^{-1})RemarksRef.
SnTe2\mathrm{SnTe}^{2}3000.0415t=20 nmt = 20\,\mathrm{nm}; (001) plane[21]
SnTe2\mathrm{SnTe}^{2}100.022t=20 nmt = 20\,\mathrm{nm}; (001) plane[21]
WTe22\mathrm{WTe}_{2}^{2}244t=35 nmt = 35\,\mathrm{nm}[22]
Mn(Bi0.7Sb0.3)4Te73\mathrm{Mn}(\mathrm{Bi}_{0.7}\mathrm{Sb}_{0.3})_{4}\mathrm{Te}_{7}^{3}40.215t=20 nmt = 20\,\mathrm{nm}[23]
Mn(Bi0.7Sb0.3)4Te73\mathrm{Mn}(\mathrm{Bi}_{0.7}\mathrm{Sb}_{0.3})_{4}\mathrm{Te}_{7}^{3}2000.015t=20 nmt = 20\,\mathrm{nm}[23]
MnBi2Te42\mathrm{MnBi}_{2}\mathrm{Te}_{4}^{2}303×10−43 \times10^{-4}7 SL[24]
MnBi2Te42\mathrm{MnBi}_{2}\mathrm{Te}_{4}^{2}3001.1×10−41.1 \times10^{-4}7 SL[24]
MnBi2Te42\mathrm{MnBi}_{2}\mathrm{Te}_{4}^{2}3151.06×10−41.06 \times10^{-4}7 SL[24]
MnBi2Te42\mathrm{MnBi}_{2}\mathrm{Te}_{4}^{2}100.028Gated[24]
MnBi2Te42\mathrm{MnBi}_{2}\mathrm{Te}_{4}^{2}3256.2×10−56.2 \times10^{-5}14 SL[24]
Te2\mathrm{Te}^{2}43×1033 \times10^{3}Gated[25]
Te2\mathrm{Te}^{2}205.3×1035.3 \times10^{3}28 nm[25]
Cd3As22\mathrm{Cd}_{3}\mathrm{As}_{2}^{2}2700.13t=95 nmt = 95\,\mathrm{nm}; at zero field; Gated[26]
TaIrTe42\mathrm{TaIrTe}_{4}^{2}201.1t=8 nmt = 8\,\mathrm{nm}; at zero field[27]
TaIrTe42\mathrm{TaIrTe}_{4}^{2}2500.015t=8 nmt = 8\,\mathrm{nm}; at zero field[27]
WSe22\mathrm{WSe}_{2}^{2}231Twisted bilayer; Gated[28]
WSe22\mathrm{WSe}_{2}^{2}2310.2Twisted bilayer; Gated[28]
6R-TaS22\mathrm{6R\text{-}TaS}_{2}^{2}3000.11[29]
Fe5GeTe2/AlxO32\mathrm{Fe}_{5}\mathrm{GeTe}_{2}/\mathrm{Al}_{x}\mathrm{O}_{3}^{2}52.93[30]
Fe5GeTe2/AlxO32\mathrm{Fe}_{5}\mathrm{GeTe}_{2}/\mathrm{Al}_{x}\mathrm{O}_{3}^{2}800.67[30]
TaIrTe41\mathrm{TaIrTe}_{4}^{1}4172t=88.2 nmt = 88.2\,\mathrm{nm}; in-plane d.c. electric field[31]
Td−WTe21T_d\mathrm{-WTe}_{2}^{1}101.13 layers[32]
CdGeAs22\mathrm{CdGeAs}_{2}^{2}30085[33]
CdGeAs22\mathrm{CdGeAs}_{2}^{2}316100[33]
CdGeAs22\mathrm{CdGeAs}_{2}^{2}35079[33]
Bi2Te31\mathrm{Bi}_{2}\mathrm{Te}_{3}^{1}3000.00630 nm[34]
Bi2Te32\mathrm{Bi}_{2}\mathrm{Te}_{3}^{2}3000.006450 nm[34]
Bi2Te31\mathrm{Bi}_{2}\mathrm{Te}_{3}^{1}2000.01230 nm[34]
Bi2Te32\mathrm{Bi}_{2}\mathrm{Te}_{3}^{2}20.0430 nm[34]
MoS21\mathrm{MoS}_{2}^{1}1020[35]
Graphene2^{2}26×1036 \times10^{3}Bernal bilayer; Gated[36]

Table S1 (continued).

Notes: 1, directly reported in text or tables; 2, extracted from figures or calculated using reported parameters and stated dimensions; 3, calculated using dimensions estimated from micrographs or scale bars.

References

  1. [1]Z. Z. Du, H.-Z. Lu, and X. C. Xie, Nat. Rev. Phys. 3, 744 (2021).
  2. [2]C. Ortix, Adv. Quantum Technol. 4, 2100056 (2021).
  3. [3]A. Bandyopadhyay, N. B. Joseph, and A. Narayan, Mater. Today Electron. 8, 100101 (2024).
  4. [4]M. Suárez-Rodríguez, F. de Juan, I. Souza, M. Gobbi, F. Casanova, and L. E. Hueso, Nat. Mater. 24, 1005 (2025).
  5. [5]S. Li, C. M. Wang, Z. Z. Du, F. Qin, H.-Z. Lu, and X. C. Xie, npj Quantum Mater. 6, 96 (2021).
  6. [6]S. Wang, W. Niu, and Y.-W. Fang, Microstructures 5, 2025060 (2025).
  7. [7]Z.-F. Wu, M.-Z. Huang, and N. Wang, Acta Phys. Sin. 72, 237301 (2023).
  8. [8]M. Wu, M. Zhang, Z. Zhang, Z. Li, N. Shi, Y. Yu, and H. Li, ACS Appl. Mater. Interfaces 18, 170 (2026).
  9. [9]I. Sodemann and L. Fu, Phys. Rev. Lett. 115, 216806 (2015).
  10. [10]Q. Ma, S.-Y. Xu, H. Shen, D. MacNeill, V. Fatemi, T.-R. Chang, A. M. Mier Valdivia, S. Wu, Z. Du, C.-H. Hsu, S. Fang, Q. D. Gibson, K. Watanabe, T. Taniguchi, R. J. Cava, E. Kaxiras, H.-Z. Lu, H. Lin, L. Fu, N. Gedik, and P. Jarillo-Herrero, Nature 565, 337 (2019).
  11. [11]K. Kang, T. Li, E. Sohn, J. Shan, and K. F. Mak, Nat. Mater. 18, 324 (2019).
  12. [12]Z. Z. Du, C. M. Wang, S. Li, H.-Z. Lu, and X. C. Xie, Nat. Commun. 10, 3047 (2019).
  13. [13]A. Gao, Y.-F. Liu, J.-X. Qiu, B. Ghosh, T. V. Trevisan, Y. Onishi, C. Hu, T. Qian, H.-J. Tien, S.-W. Chen, M. Huang, D. Bérubé, H. Li, C. Tzschaschel, T. Dinh, Z. Sun, S.-C. Ho, S.-W. Lien, B. Singh, K. Watanabe, T. Taniguchi, D. C. Bell, H. Lin, T.-R. Chang, C. R. Du, A. Bansil, L. Fu, N. Ni, P. P. Orth, Q. Ma, and S.-Y. Xu, Science 381, 181 (2023).
  14. [14]N. Wang, D. Kaplan, Z. Zhang, T. Holder, N. Cao, A. Wang, X. Zhou, F. Zhou, Z. Jiang, C. Zhang, S. Ru, H. Cai, K. Watanabe, T. Taniguchi, B. Yan, and W. Gao, Nature 621, 487 (2023).
  15. [15]A. Tiwari, F. Chen, S. Zhong, E. Drueke, J. Koo, A. Kaczmarek, C. Xiao, J. Gao, X. Luo, Q. Niu, Y. Sun, B. Yan, L. Zhao, and A. W. Tsen, Nat. Commun. 12, 2049 (2021).
  16. [16]D. Kumar, C.-H. Hsu, R. Sharma, T.-R. Chang, P. Yu, J. Wang, G. Eda, G. Liang, and H. Yang, Nat. Nanotech. 16, 421 (2021).
  17. [17]P. He, H. Isobe, D. Zhu, C.-H. Hsu, L. Fu, and H. Yang, Nat. Commun. 12, 698 (2021).
  18. [18]E. Lesne, Y. G. Sağlam, R. Battimello, M. T. Mercaldo, T. C. van Thiel, U. Filippozzi, C. Noce, M. Cuoco, G. A. Steele, C. Ortix, and A. D. Caviglia, Nat. Mater. 22, 576 (2023).
  19. [19]H. Zhang, D. Tian, X. Chen, W. Qi, L. Chen, M. Li, Y. Bai, J. Zhang, F. Han, H. Yang, Y. Chen, Y. Chen, J. Wu, Y. Xu, F. Hu, B. Shen, J. Sun, and W. Zhao, Adv. Mater. 38, e16347 (2026).
  20. [20]S. Dash and P. Padhan, Adv. Funct. Mater. 35, e07191 (2025).
  21. [21]B. Cheng, Y. Gao, Z. Zheng, S. Chen, Z. Liu, L. Zhang, Q. Zhu, H. Li, L. Li, and C. Zeng, Nat. Commun. 15, 5513 (2024).
  22. [22]H. Liu, T. Y. Lim, S. Tian, J. Zhai, D. Xi- ang, T. Liu, T. Chang, P. He, and J. Shen, Adv. Electron. Mater. 11, 2400648 (2025).
  23. [23]S. H. Lee, T. Iwaya, K. Nakayama, T. Y. Lim, L. Min, J. He, Y. Wang, V. Gopalan, Z. Xie, X. Pan, Y. P. Chen, T. Chang, H. Lin, L. Fu, K. Segawa, T. Sato, and Z. Mao, Adv. Mater. 38, e14217 (2026).
  24. [24]G. Kim, J. Bahng, J. Jeong, W. Sakong, T. Lee, D. Lee, Y. Kim, H. Rho, and S. C. Lim, Nano Lett. 24, 10820 (2024).
  25. [25]R. Alam, D. Dutta, M. Sariket, N. Kumar, A. Agarwal, and A. N. Pal, Adv. Funct. Mater. 35, 2418643 (2025).
  26. [26]S. Wu, W. Tian, R. Li, Z. Han, X. Zhou, S. Huang, P. Li, P. Song, and X. Zhao, Matter 8, 102153 (2025).
  27. [27]D. Zhao, Z. Li, J. Sun, Y. Zhao, T. Wang, L. Qi, W. Tang, S. Zhang, A. Ramiere, H. Jin, H. Liang, K. P. Loh, and Y.-J. Zeng, ACS Nano 20, 16325 (2026).
  28. [28]Z. Hu, L. Zhang, A. Chakraborty, G. D’Olimpio, J. Fujii, A. Ge, Y. Zhou, C. Liu, A. Agarwal, I. Vobornik, D. Farias, C. Kuo, C. S. Lue, A. Politano, S. Wang, W. Hu, X. Chen, W. Lu, and L. Wang, Adv. Mater. 35, 2209557 (2023).
  29. [29]T. Xi, H. Jiang, J. Li, Y. He, Y. Gu, C. Fox, L. Primeau, Y. Mao, J. Rollins, T. Taniguchi, K. Watanabe, D. van der Weide, D. Rhodes, Y. Zhang, Y. Wang, and J. Xiao, Nat. Electron. 8, 578 (2025).
  30. [30]L. Min, H. Tan, Z. Xie, L. Miao, R. Zhang, S. H. Lee, V. Gopalan, C.-X. Liu, N. Alem, B. Yan, and Z. Mao, Nat. Commun. 14, 364 (2023).
  31. [31]S. H. Lee, Y. Zhao, N. Gluscevich, H. Yi, Z. Morgan, H. Cao, J. Koo, Y. Wang, J. He, V. Gopalan, Y. Wang, W. Xie, C. Chang, Q. Zhang, B. Yan, and Z. Mao, Adv. Funct. Mater. 35, 2424231 (2025).
  32. [32]P. Makushko, S. Kovalev, Y. Zabila, I. Ilyakov, A. Ponomaryov, A. Arshad, G. L. Prajapati, T. V. A. G. de Oliveira, J.-C. Deinert, P. Chekhonin, I. Veremchuk, T. Kosub, Y. Skourski, F. Ganss, D. Makarov, and C. Ortix, Nat. Electron. 7, 207 (2024).
  33. [33]Z. Liu, J. Shi, J. Cao, Z. Ma, Z. Yang, Y. Cui, L. Wang, Y. Dai, M. Chen, P. Wang, Y. Xie, F. Chen, Y. Shi, C. Xiao, S. A. Yang, B. Cheng, S. Liang, and F. Miao, Adv. Funct. Mater. 35, 2416204 (2025).
  34. [34]J. Chen, S. Kang, G. Zhan, S. Gu, J. Jiang, W. Wang, G. Hu, Y. Liu, M. Liu, and L. Pan, Appl. Phys. Express 18, 046502 (2025).
  35. [35]D. Wu, Q. M. Liu, S. L. Chen, G. Y. Zhong, J. Su, L. Y. Shi, L. Tong, C. Xu, P. Gao, and N. L. Wang, Phys. Rev. Mater. 3, 024002 (2019).
  36. [36]Q. M. Liu, D. Wu, T. Wu, S. Han, Y. Peng, Z. Yuan, Y. Cheng, B. Li, T. Hu, L. Yue, S. Xu, R. Ding, M. Lu, R. Li, S. Zhang, B. Lv, A. Zong, Y. Su, N. Gedik, Z. Yin, T. Dong, and N. Wang, Nat. Commun. 15, 8937 (2024).
  37. [37]B. Q. Lv, A. Zong, D. Wu, A. V. Rozhkov, B. V. Fine, S.-D. Chen, M. Hashimoto, D.-H. Lu, M. Li, Y.-B. Huang, J. P. C. Ruff, D. A. Walko, Z. H. Chen, I. Hwang, Y. Su, X. Shen, X. Wang, F. Han, H. C. Po, Y. Wang, P. Jarillo-Herrero, X. Wang, H. Zhou, C.-J. Sun, H. Wen, Z.-X. Shen, N. L. Wang, and N. Gedik, Phys. Rev. Lett. 128, 036401 (2022).
  38. [38]R. Venturini, M. Rupnik, J. Gašperlin, J. Lipič, P. Šutar, Y. Vaskivskyi, F. Ščepanović, D. Grabnar, D. Golež, and D. Mihailovic, Nat. Commun. 17, 8360 (2026).
  39. [39]B. Lv, Y. Su, A. Zong, K. Morey, B. T. Fichera, Q. Liu, D. Wu, Y. Ma, D. Zhang, F. Zhou, M. Hashimoto, D.-H. Lu, D. A. Walko, H. Wen, J. Li, S. Sarker, J. P. C. Ruff, N. Wang, and N. Gedik, Nat. Commun. 17, 7123 (2026).
  40. [40]B. Q. Lv, Y. Su, A. Zong, Q. Liu, D. Wu, N. F. Q. Yuan, Z. Nie, J. Li, S. Sarker, S. Meng, J. P. C. Ruff, N. L. Wang, and N. Gedik, Nat. Mater. 25, 240 (2026).
  41. [41]Q. Q. Zhang, Y. Shi, K. Y. Zhai, W. X. Zhao, X. Du, J. S. Zhou, X. Gu, R. Z. Xu, Y. D. Li, Y. F. Guo, Z. K. Liu, C. Chen, S.-K. Mo, T. K. Kim, C. Cacho, J. W. Yu, W. Li, Y. L. Chen, J.-H. Chu, and L. X. Yang, Phys. Rev. B 107, 115141 (2023).
  42. [42]W.-H. Dong, W. Y. Huang, Y. Xu, and P. Tang, “Multiple layer-selective polar charge density waves in EuTe4,” (2026), 2602.09855 [cond-mat.mtrl-sci].
  43. [43]S. Dzsaber, X. Yan, M. Taupin, G. Eguchi, A. Prokofiev, T. Shiroka, P. Blaha, O. Rubel, S. E. Grefe, H.-H. Lai, Q. Si, and S. Paschen, Proc. Natl. Acad. Sci. U.S.A. 118, e2013386118 (2021).
  44. [44]M. Huang, Z. Wu, J. Hu, X. Cai, E. Li, L. An, X. Feng, Z. Ye, N. Lin, K. T. Law, and N. Wang, Natl. Sci. Rev. 10, nwac232 (2023).
  45. [45]J. Duan, Y. Yan, Y. Jian, Y. Gao, H. Peng, J. Zhong, Q. Feng, J. Mao, and Y. Yao, Phys. Rev. Lett. 129, 186801 (2022).
  46. [46]M.-S. Qin, P.-F. Zhu, X.-G. Ye, W.-Z. Xu, Z.-H. Song, J. Liang, K. Liu, and Z.-M. Liao, Chin. Phys. Lett. 38, 017301 (2021).
  47. [47]X.-G. Ye, H. Liu, P.-F. Zhu, W.-Z. Xu, S. A. Yang, N. Shang, K. Liu, and Z.-M. Liao, Phys. Rev. Lett. 130, 016301 (2023).
  48. [48]Y. M. Itahashi, T. Ideue, S. Hoshino, C. Goto, H. Namiki, T. Sasagawa, and Y. Iwasa, Nat. Commun. 13, 1659 (2022).
  49. [49]N. Wang, J.-Y. You, A. Wang, X. Zhou, Z. Zhang, S. Lai, Y.-P. Feng, H. Lin, G. Chang, and W.-b. Gao, Natl. Sci. Rev. 11, nwad103 (2024).
  50. [50]X. F. Lu, C.-P. Zhang, N. Wang, D. Zhao, X. Zhou, W. Gao, X. H. Chen, K. T. Law, and K. P. Loh, Nat. Commun. 15, 245 (2024).
  51. [51]F. Hu, P. Zhao, L. Yang, S. Zhao, J. Lei, W. Li, J. Lai, Z. Yu, H. Park, C. Wong, R. Sharma, G. Eda, S. A. Yang, X. Xu, F. Wang, and H. Yang, Nat. Nanotechnol. 20, 1588 (2025).
  52. [52]H. Li, M. Li, R.-C. Xiao, W. Liu, L. Wu, W. Gan, H. Han, X. Tang, C. Zhang, and J. Wang, Appl. Phys. Lett. 123, 163102 (2023).
  53. [53]S. Wang, X. Li, H. Zhang, B. Chen, H. Xie, C. Li, F. Fei, S. Zhang, and F. Song, Appl. Phys. Lett. 124, 153102 (2024).
  54. [54]S. Liu, R. Burgos, E. Zhang, N. Wang, X.-B. Qiang, C. Li, Q. Zhang, Z. Z. Du, R. Zheng, J. Chen, Q.-H. Xu, K. Leng, W. Gao, F. Xiu, D. Culcer, and K. P. Loh, Commun. Phys. 7, 413 (2024).
  55. [55]T.-Y. Zhao, A.-Q. Wang, Z.-T. Zhang, Z.-Y. Cao, X.-Y. Liu, and Z.-M. Liao, Phys. Rev. Lett. 135, 026601 (2025).
  56. [56]A.-Q. Wang, Di. Li, T.-Y. Zhao, X.-Y. Liu, J. Zhang, X. Liao, Q. Yin, Z.-C. Pan, P. Yu, and Z.-M. Liao, Phys. Rev. B 110, 155434 (2024).
  57. [57]Z.-Y. Cao, A.-Q. Wang, X.-Y. Liu, T.-Y. Zhao, D. Yu, and Z.-M. Liao, Phys. Rev. B 111, 125407 (2025).
  58. [58]J. Zhong, H. Peng, X. Wang, Q. Feng, Y. Hu, Q. Li, Y. Li, W. Jiang, Z. Wang, and J. Duan, Phys. Rev. Applied 21, 024044 (2024).
  59. [59]J. Yang, L. Wei, Y. Li, L. Chen, W. Niu, S. Wang, F. Li, P. Liu, S. Zhou, and Y. Pu, Appl. Phys. Lett. 127, 033102 (2025).
  60. [60]S. Wu, Y.-F. Liu, S. Y. Kwon, T. Y. Lim, S. Yeom, A. Gao, H. Li, R. R. Unocic, A. M. Varghese, E. Lopez-Torres, K. Xiao, T.-R. Chang, M. Yoon, S.-Y. Xu, and X. Li, Nano Lett. 25, 17475 (2025).
  61. [61]Q. Shen, J. Chen, B. Rong, Y. Rong, H. Chen, X. Duan, T. Zhao, D. Guan, S. Wang, Y. Li, H. Zheng, X. Liu, X. Qiu, J. Chen, L. Cong, T. Li, R. Zhong, C. Liu, Y. Yang, L. Liu, and J. Jia, Adv. Funct. Mater. 35, e06694 (2025).
  62. [62]F. Zhou, Y. Du, T. Wang, H. Zhang, J. Li, W. Qi, Y. Yu, F. Fei, and F. Song, Mater. Futures 5, 025302 (2026).
  63. [63]H. Chen, Q. Hu, P. Boval, and K. P. Loh, Nat. Commun. 17, 8856 (2026).

Paper details

vlnc:W_pm2sfxjh
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