Abstract
The ultraquantum limit is achieved when a magnetic field confines an electron gas in its lowest spinpolarised Landau level. Here we show that in this limit, electron doped ZrTe_{5} shows a metalinsulator transition followed by a sign change of the Hall and Seebeck effects at low temperature. We attribute this transition to a magnetic freezeout of charge carriers on the ionized impurities. The reduction of the charge carrier density gives way to an anomalous Hall response of the spinpolarised electrons. This behavior, at odds with the usual magnetic freezeout scenario, occurs in this Dirac metal because of its tiny Fermi energy, extremely narrow band gap and a large gfactor. We discuss the different possible sources (intrinsic or extrinsic) for this anomalous Hall contribution.
Similar content being viewed by others
Introduction
In the presence of a magnetic field, the electronic spectrum of a threedimensional electron gas (3DEG) is quantized into Landau levels. When all the charge carriers are confined in the lowest Landau level—the socalled quantum limit—the kinetic energy of electrons is quenched in the directions transverse to the field. This favors the emergence of electronic instabilities, either driven by the electronelectron or electronimpurity interactions^{1,2,3,4}. So far, the behavior of 3DEGs beyond their quantum limit has been explored in a limited number of low carrier density systems. Yet, different instabilities have been detected, such as a thermodynamic phase transition in graphite^{5,6,7,8}, a valley depopulation phase in bismuth^{9,10}, and a metalinsulator transition (MIT) in narrowgap doped semiconductors InSb^{11} and InAs^{12,13}. The latter occurs when charge carriers are confined in the lowest spinpolarised Landau level—the ultraquantum limit. This transition is generally attributed to the magnetic freezeout effect where electrons are frozen on ionized impurities^{4,14}.
Lately, lowdoped Dirac and Weyl materials with remarkable fieldinduced properties were discovered^{15,16,17,18,19}. Of particular interest is the case of ZrTe_{5}. The entrance into its quantum limit regime is marked by quasi quantized Hall resistivity (ρ_{xy})^{18} and thermoelectrical Hall conductivity (α_{xy})^{20,21}, followed by a higher magnetic field transition^{18,22}. This phase transition has initially been attributed to the formation of a charge density wave (CDW)^{18,22,23}. Such interpretation has been questioned because of the absence of thermodynamic evidence^{24,25}, expected for a CDW transition. Furthermore, ZrTe_{5} displays a large anomalous Hall effect (AHE), even though it is a nonmagnetic material^{26,27,28,29,30}.
Here we report electrical, thermoelectrical and optical conductivity measurements over a large range of doping, magnetic field, and temperature in electrondoped ZrTe_{5}. This allows us to track the Fermi surface evolution of ZrTe_{5} and explain the nature of this phase transition, as well as its links with the observed AHE. We show that the onset of the fieldinduced transition can be ascribed to the magnetic freezeout effect. In contrast with usually reported results, we show that the freezeout regime of ZrTe_{5} is characterized by a sign change of the Hall and thermoelectric effects, followed by a saturating Hall conductivity. Our results show that the magnetic freezeout effect differs in this Dirac material as a consequence of the tiny band gap and large gfactor of ZrTe_{5}, that favor both an extrinsic and an intrinsic AHE of the spinpolarised charge carriers.
Results
Fermi surface of ZrTe_{5}
Figure 1 a shows the temperature dependence of the resistivity (ρ_{xx}) for four batches, labelled S_{1−4} respectively. Samples from the same batch are labelled by distinct subscript letters (see Supplementary Note 1). At room temperature, ρ_{xx} ≈ 0.7 mΩ.cm. With decreasing temperature, ρ_{xx} peaks at a temperature around which the Hall effect (ρ_{xy}) changes sign, which is around 150 K for S_{3b} sample (see Fig. 1b). Both shift to lower temperature as the carrier density decreases. These effects have been tracked by laser angleresolved photoemission spectroscopy and attributed to a temperatureinduced phase transition where the Fermi energy shifts from the top of the valence band to the bottom of the conduction band as the temperature decreases^{31}.
At low temperature the Fermi energy is located in the conduction band. Fig. 1c, d show the quantum oscillations for samples from batches S_{1}, S_{2} and S_{3} for a magnetic field (B) parallel to the baxis of the orthorhombic unit cell. The angular dependence of the quantum oscillation frequency are well fitted by an anisotropic ellipsoid Fermi surface elongated along the baxis, and in good agreement with previous measurements^{24,32,33} (see Fig. 1g). Our doping study reveals that the ellipsoid anisotropy increases as the system is less doped, see Fig. 1f. In our lowest doped samples the ratio of the Fermi momentum (k_{F}) along the a and baxis reach 0.06 implying a mass anisotropy ratio of \(\frac{{m}_{b}^{* }}{{m}_{a}^{* }}\simeq\) 250, where \({m}_{a,b}^{* }\) are the band mass along the a and b axis. This large mass anisotropy ratio is comparable to the one of Dirac electrons of bismuth^{34}. This Fermi surface mapping allows us to accurately determine the Fermi sea carrier densities, n_{SdH}, which agree well with n_{H} (see Supplementary Note 2). Remarkably S_{1} samples have a Hall mobility, μ_{H}, as large as 9.7 × 10^{5} V ⋅ cm^{−2} ⋅ s^{−1} and the last quantum oscillation occurs at a small field of B_{QL}(S_{1}) = 0.3 T for B∥b. Given the large gfactor, g^{*} ≈ 20–30^{22,35}, this last oscillation corresponds to the depopulation of the (0, +) Landau level. Above it the highly mobile electrons are all confined into the lowest spinpolarised (0, −) Landau level.
Field induced transition in the ultraquantum limit of ZrTe_{5}
Figure 2 shows the field dependence of ρ_{xx} beyond the ultraquantum limit of S_{1}, S_{2} and S_{3} samples. In the lowest doped samples (S_{1}) ρ_{xx} increases by more than two orders of magnitude and saturates above ≈7 T. This large magnetoresistance vanishes as the temperature increases (see Supplementary Fig. 2), for T > 5 K and up to 50 T. A close inspection of the low temperature behavior reveals a light metallic phase above B_{QL} (see Fig. 2a, b) which ends at a crossing point at B_{c} = 3.2 T above which an insulating state is observed up to 50 T. Following^{18} we take this crossing point as the onset of the field induced metalinsulator transition. As the carrier density increases, both the position of B_{QL} and B_{c} increase (see Fig. 2b, c)). At the highest doping (samples S_{3}) the amplitude of the magnetoresistance has decreased and the transition is only marked by a modest increase by a factor of two of ρ_{xx} at ≃ 30 T, indicating that the transition smears with increasing doping (see Fig. 2d). Figure 2e shows the doping evolution of B_{QL} and B_{c} which are in good agreement with previous works^{18,22,24}. For an isotropic 3D Dirac material \({B}_{QL}=\hslash /e{(\sqrt{(2)}{\pi }^{2}n)}^{2/3}\) (see i.e^{36}) with \(n=3{\pi }^{2}{k}_{F}^{3}\). In the B∥b configuration k_{F} = \(\sqrt{({k}_{F,a}{k}_{F,c})}\) can be evaluated from the frequency of quantum oscillations. The deduced B_{QL} is shown by the red line in Fig. 2e and provides an excellent agreement with the detected B_{QL}. As function of the total carrier density of the ellipsoid (n_{SdH}) \({B}_{QL}=\hslash /e{(\sqrt{(2{A}_{1}{A}_{2})}{\pi }^{2}{n}_{SdH})}^{2/3}\) where A_{1} and A_{2} are the anisotropic Fermi momentum ratios between the a and baxis, and between the c and baxis.
The doping evolution of B_{c} is a clue to the nature of this transition. So far it has been attributed to the formation of a charge density wave (CDW) along the magnetic field^{18,22,23}. Such an instability is favored by the onedimensional nature of the electronic spectrum along the magnetic field, which provides a suitable (2k_{F}) nesting vector in the (0, −) Landau level. In this picture, predicted long ago^{1}, the transition is of second order and is expected to vanish as the temperature increases. The absence of temperature dependence of B_{c} and the absence of thermodynamic signature^{24,25} invite us to consider another interpretation.
In the CDW picture, the instability is driven by the electronelectron^{1,18,22} or electronphonon interaction^{23} and the interaction between electrons and the ionized impurities is neglected. However, in a doped semiconductor, the conduction band electrons are derived from uncompensated donors. Tellurium vacancies have been identified as the main source of impurities in ZrTe_{5} flux grown samples^{37,38}. According to the Mott criterion^{39,40} a semiconductor becomes metallic when the density of its carriers, n, exceeds a threshold set by its effective Bohr radius, a_{B} = 4πεℏ/m^{*}e^{2} (where m^{*} is the effective mass of the carrier, ε is the dielectric constant of the semiconductor): n^{1/3}a_{B} ≃ 0.3. In presence of a magnetic field the inplane electronic wave extension shrinks with increasing magnetic field. When B > B_{QL}, the inplane Bohr radius is equal to a_{B,⊥} = 2ℓ_{B} with \({\ell }_{B}=\sqrt{(\frac{\hslash }{eB})}\)^{4,41}. Along the magnetic field direction, the characteristic spatial extension is \({a}_{B,\parallel }=\frac{{a}_{B,z}}{\log (\gamma )}\), where \(\gamma ={(\frac{{a}_{B,c}}{{l}_{B}})}^{2}\) with a_{B,z}=\(\frac{\varepsilon }{{m}_{z}^{* }}{a}_{B,0}\) and a_{B,c}=\(\frac{\varepsilon }{{m}_{c}^{* }}{a}_{B,0}\), where \({m}_{z,c}^{* }\) are the mass along and perpendicular to the magnetic field in units of m_{0}, and a_{B,0} the bare Bohr radius. A MIT transition is thus expected to occur when the overlap between the wave functions of electrons is sufficiently decreased^{11,14} i.e. when:
This MIT is thus a Mott transition assisted by the magnetic field where the metal is turned into an insulator due to the freezing of electrons on the ionized donors by the magnetic field, the socalled magnetic freezeout effect. According to Eq. (1), \(n\propto {B}_{c}/\log ({B}_{c})\) and B_{c} is slightly sublinear in n and evolves almost parallel to B_{QL}. In order to test this scenario quantitatively, one has to determine the threshold of the transition from Eq. (1), which requires knowing ε and \({m}_{z/c}^{* }\). Temperature dependence of the quantum oscillations gives access to \({m}_{z}^{* }\,\) ≈ 2m_{0} and \({m}_{c}^{* }\,\) ≈ 0.02m_{0} for B∥b, while the optical reflectivity measurements give access to ε. Figure 3 shows ε versus temperature for two samples of batches S_{1} and S_{3}. ε is as large as 200400ε_{0} in ZrTe_{5} (see Supplementary Note 5). The deduced onset from Eq. (1) is shown in dashed black lines in Fig. 2e for ε = 200 and 400, capturing well the doping evolution of B_{c}. We thus attribute the transition detected in the ultraquantum limit of ZrTe_{5} to the magnetic freezeout effect.
It is worth noticing that a large contribution to ε comes from interband electronic transitions resulting in ε_{∞} > 100. This result also clarifies why one can detect highly mobile carriers even down to densities as low as 10^{13} cm^{−3}^{29}. Due to the light inplane carrier mass and large dielectric constant, one expects the threshold of the MIT at zero magnetic field to be below ≈10^{12} cm^{−3}.
Discussion
In InSb (n_{H} = 2–5 × 10^{15} cm^{−3})^{11}, a large drop of the carrier density comes with an activated insulating behavior. In contrast with that usual freezeout scenario, we find in ZrTe5 a rather soft insulating behavior, where ρ_{xx} saturates at the lowest temperature. Measurements of the Hall effect and thermoelectrical properties at subkelvin temperatures shown in Fig. 4a–c reveal an unexpected field scale, thus confirming that the freezeout regime of ZrTe_{5} differs from the usual case. Above 7 T, ρ_{xy} and the Seebeck effect (S_{xx} = \(\frac{{E}_{x}}{{{{\Delta }}}_{x}t}\)) change signs and saturate from 10 T up to 50 T for ρ_{xy} (see Supplementary Notes 3 and 4). The field induced sign changes of ρ_{xy} and S_{xx} are reminiscent of the sign change in temperature.
The temperature dependence of S_{xx}/T for B = 0, 6 and 12 T (shown in Fig. 4e) enables us to quantify the variations of charge carrier density as a function of the magnetic field. At B = 0 T, S_{xx}/T = − 5.5 μV.K^{−2}, which is in quantitative agreement with the expected value for the diffusive response of a degenerate semiconductor: \({S}_{xx}/T=\frac{{\pi }^{2}}{2}\frac{{k}_{B}}{e{T}_{F}}=5\) μV.K^{−2} for T_{F} ≈ 80 K deduced from quantum oscillation measurements. At B = 12 T S_{xx}/T saturates, at low temperature, to ≃ +20 μV.K^{−2}, a value which is four times larger than at zero magnetic field, pointing to a reduction of the charge carrier density by only a factor of eight. The partial freezeout of the charge carriers is the source of the saturating ρ_{xx}. We now discuss the specificity of ZrTe_{5} that leads to this peculiar freezeout regime.
In the kspace, the magneticfreeze out transition corresponds to a transfer of electrons from the lowest Landau level (0, −) to a shallow band, see inset of Fig. 2e, formed by the localized electrons^{4}. This theory does not fully apply to ZrTe_{5} for two reasons. First, it applies to large gap systems with no potential spatial fluctuations, and ZrTe_{5} has only a band gap of 6 meV^{42}, which is fifty times smaller than that of narrow gap semiconductors such as InSb or InAs. Second, the Fermi surface of ZrTe_{5} is highly anisotropic. The same critical field is thus reached for a carrier density that is fifty times larger in ZrTe_{5} than in isotropic Fermi surface materials, like InSb or InAs. The large Bohr radius and the relatively higher density of ZrTe_{5} will therefore inevitably broaden the density of states, set by: \({{\Gamma }}=2\sqrt{\pi }\frac{{e}^{2}}{\epsilon {r}_{s}}{({N}_{i}{r}_{s}^{3})}^{\frac{1}{2}}\) where \({r}_{s}\propto \sqrt{(\frac{{a}_{B}}{n\frac{1}{3}})}\) is the screening radius and N_{i} is an estimate of the impurity concentration^{43}. Assuming that n ≈ N_{i}, we estimate Γ ≈ 6 meV in S_{1} samples.
In contrast with other narrowgap semiconductors where Γ < < E_{F} < < Δ, the magnetic freezeout occurs in ZrTe_{5} where Γ ≈ E_{F} ≈ Δ. In this limit, the shallow band of width Γ will overlap the LLL of the conduction band, and eventually the valence band giving rise to a finite residual electron and hole charge carriers at low temperature as sketched on Fig. 2e. As a function of doping, Γ increases the smearing of the transition (Fig. 2). The convergence of the three energy scales Γ, E_{F} and Δ is one source of the partial reduction of charge carrier density detected in ρ_{xx}, S_{xx} and of the sign change of ρ_{xy}. This finite residual charge carrier should give rise to a linear Hall effect, contrasting with the saturating ρ_{xy} (and σ_{xy}), which is typical of an anomalous response. We discuss this anomalous contribution in the last section.
Several studies have reported an AHE in ZrTe_{5}^{26,27,28,29}. In this case, the Hall conductivity is the sum of two contributions: \({\sigma }_{xy}=\frac{ne}{B}+{\sigma }_{xy}^{A}\) where the first and second terms are the orbital conductivity and the anomalous Hall conductivity, respectively. At high enough magnetic field, \({\sigma }_{xy}^{A}\) becomes dominant, setting the amplitude and the sign of ρ_{xy}. So far, \({\sigma }_{xy}^{A}\) has been attributed to the presence of a nonzero Berry curvature—an intrinsic effect—either due to the Weyl nodes in the band structure^{26}, or to the spinsplit massive Dirac bands with non zero Berry curvature^{28,29}. In the latter case, \({\sigma }_{xy}^{A}\) scales with the carrier density, and its amplitude is expected to be +1 (Ω.cm)^{−1} for n_{H} = 2 × 10^{16} cm^{−3}^{28}, which is of the same order of magnitude as our results. Skew and side jump scattering are another source of AHE in non magnetic semiconductors^{44,45}. Deep in the freezeout regime of low doped InSb (n_{H} ≈ 10^{14} cm^{−3}), a sign change of the Hall effect has been observed and attributed to skew scattering^{46}. In contrast with dilute ferromagnetic alloys, where the asymmetric electron scattering is due to the spinorbit coupling at the impurity sites, here it is caused by the spinpolarised electron scattering by ionized impurities. Its amplitude is given by \({\sigma }_{xy}^{S}\)= N_{S}e\(\frac{{g}^{* }{\mu }_{B}}{{E}_{1}}\), where E_{1}=\(\frac{{\epsilon }_{G}({\epsilon }_{G}+{{\Delta }})}{2{\epsilon }_{G}+{{\Delta }}}\) with ϵ_{G} the band gap and Δ the spinorbit splitting of the valence band. N_{S} = N_{A} + n is the density of positively charged scattering centers with N_{A} the density of acceptors^{46}. Note that \({\sigma }_{xy}^{S}\) induces a sign change of the Hall conductivity and is only set by intrinsic parameters and by N_{S}. Assuming N_{S} ≈ n_{H}(B = 0), and taking g^{*} ≈ 20^{22,35} and E_{1} = ϵ_{G} = 6 meV (ϵ_{G} < < Δ), we find that \({\sigma }_{xy}^{S}\) ≃ +1 (Ω. cm)^{−1}, which is similar to the intrinsic contribution. Remarkably, it is four orders of magnitude larger than what has been observed in low doped InSb^{46}, due to the tiny gap and a (relatively) larger carrier density in ZrTe_{5}.
Therefore, the AHE contribution can induce a sign change of ρ_{xy} in electron doped ZrTe_{5}. It is accompanied by a peak in S_{xx}/T (see Fig. 4c), S_{xy}/T (see Supplementary Figs. 3–5) and thus in α_{xy} = σ_{xx}S_{xy} + σ_{xy}S_{xx} (see Fig. 4d–f). Our result shows that the thermoelectric Hall plateau^{20,21}, observed above 5 K, collapses at low temperature. These peaks can be understood qualitatively through the Mott relation^{47} (\(\frac{\alpha }{T}=\frac{{\pi }^{2}}{3}\frac{{k}_{B}}{e}\frac{\partial \sigma (\epsilon )}{\partial \epsilon }{ }_{\epsilon = {\epsilon }_{F}}\)). This is the region where ρ_{xx} and ρ_{xy} (and thus σ_{xx} and σ_{xy}) change the most in field and temperature, so that S_{xx} and α_{xy} are the largest. The increase occurs in the vicinity of B_{c}, causing a peak in the field dependence of S_{xx} and α_{xy}, as it happens across the freezeout regime of InAs^{13}. Whether the Mott relation can quantitatively explain the amplitude of these peaks and the sign change of S_{xx} remains to be determined. This calls to extend theoretical works^{48,49,50} on the electrical and thermoelectrical response to the freezeout regime of Dirac materials such as ZrTe_{5}.
In summary, we show that the doping evolution of the onset transition detected in the ultraquantum limit of ZrTe_{5} can be ascribed to the magnetic freezeout, where electrons become bound to donors. In contrast to the usual case, the freezeout regime of ZrTe_{5} is marked by a modest reduction of the charge carrier density due to the convergence of three tiny energy scales in this Dirac material: the band gap, the slowly varying potential fluctuations and the Fermi energy. Deep in the freezeout regime, the Hall conductivity changes sign and becomes anomalous with a relatively large amplitude for this low carrier density and non magnetic material. This AHE could thus have an extrinsic origin due to skewscattering of the spinpolarised electrons by ionized impurities. Distinguishing and tuning both intrinsic and extrinsic contributions by varying the charge compensation or strain^{51} is an appealing perspective for future research. To date, the AHE of the spinpolarised electrons in the ultraquantum limit has been detected in a limited number of cases. Many Dirac materials with small gaps and large gfactors remain to be studied, in particular at higher doping where the intrinsic and extrinsic AHE are both expected to be larger.
Method
Samples and measurements description
Two sets of ZrTe_{5} samples have been used in this study. The first ones, grown by flux method where iodine served as a transport agent for the constituents, have the lowest carrier density. The second ones, grown by Chemical Vapor Transport (CVT), have the highest density. Electrical and thermal transport measurements have been measured using four point contacts. Contact resistance of a less 1 Ω has been achieved by an Argon etching, follow by the deposit of 10 nm Ti buffer layer and of 150 nm Pd layer. High magnetic field measurement has been done at LNCMIToulouse. Thermoelectrical and thermal transport measurements has been done using a standard twothermoemeters oneheater set up similar to one used in ref. ^{13}. Further experimental details can be found in Supplementary Note 1.
Data availability
All data supporting the findings of this study are available from the corresponding author B.F. upon request.
References
Celli, V. & Mermin, N. D. Ground state of an electron gas in a magnetic field. Phys. Rev. 140, A839 (1965).
Halperin, B. I. Possible states for a threedimensional electron gas in a strong magnetic field. Jpn. J. Appl. Phys 26, 1913 (1987).
MacDonald, A. H. & Bryant, G. W. Strongmagneticfield states of the pure electron plasma. Phys. Rev. Lett. 58, 515 (1987).
Yafet, Y., Keyes, R. & Adams, E. Hydrogen atom in a strong magnetic field. J. Phys. Chem. Solids 1, 137–142 (1956).
Fauqué, B. et al. Two phase transitions induced by a magnetic field in graphite. Phys. Rev. Lett. 110, 266601 (2013).
LeBoeuf, D. et al. Thermodynamic signatures of the fieldinduced states of graphite. Nat. Commun. 8, 1337 (2017).
Zhu, Z. et al. Graphite in 90 T: evidence for strongcoupling excitonic pairing. Phys. Rev. X 9, 011058 (2019).
Marcenat, C. et al. Wide critical fluctuations of the fieldinduced phase transition in graphite. Phys. Rev. Lett. 126, 106801 (2021).
Zhu, Z. et al. Emptying Dirac valleys in bismuth using high magnetic fields. Nat. Commun. 8, 15297 (2017).
Iwasa, A. et al. Thermodynamic evidence of magneticfieldinduced complete valley polarization in bismuth. Sci. Rep. 9, 1672 (2019).
Shayegan, M., Goldman, V. J. & Drew, H. D. Magneticfieldinduced localization in narrowgap semiconductors Hg_{1−x}Cd_{x}Te and InSb. Phys. Rev. B 38, 5585–5602 (1988).
Kaufman, L. A. & Neuringer, L. J. Magnetic freezeout and band tailing in nInAs. Phys. Rev. B 2, 1840–1846 (1970).
Jaoui, A. et al. Giant seebeck effect across the fieldinduced metalinsulator transition of inas. npj Quantum Mater. 5, 94 (2020).
Aronzon, B. A. & Tsidilkovskii, I. M. Magneticfieldinduced localization of electrons in fluctuation potential wells of impurities. Phys. Status Solidi (B): Basic Res. 157, 17–59 (1990).
Moll, P. J. W. et al. Magnetic torque anomaly in the quantum limit of Weyl semimetals. Nat. Commun. 7, 12492 (2016).
Ramshaw, B. et al. Quantum limit transport and destruction of the Weyl nodes in TaAs. Nat. Commun. 9, 2217 (2018).
Liang, S. et al. A gapprotected zeroHall effect state in the quantum limit of the nonsymmorphic metal KHgSb. Nat. Mater. 18, 443–447 (2019).
Tang, F. et al. Threedimensional quantum Hall effect and metalinsulator transition in ZrTe_{5}. Nature 569, 537–541 (2019).
Gooth, J. et al. Axionic chargedensity wave in the Weyl semimetal (TaSe_{4})_{2}I. Nature 575, 315–319 (2019).
Zhang, W. et al. Observation of a thermoelectric Hall plateau in the extreme quantum limit. Nat. Commun. 11, 1046 (2020).
Han, F. et al. Quantized thermoelectric Hall effect induces giant power factor in a topological semimetal. Nat. Commun. 11, 6167 (2020).
Liu, Y. et al. Zeeman splitting and dynamical mass generation in Dirac semimetal ZrTe_{5}. Nat. Commun. 7, 12516 (2016).
Qin, F. et al. Theory for the chargedensitywave mechanism of 3d quantum Hall effect. Phys. Rev. Lett. 125, 206601 (2020).
Galeski, S. et al. Origin of the quasiquantized Hall effect in ZrTe_{5}. Nat. Commun. 12, 3197 (2021).
Tian, Y., Ghassemi, N. & Ross, J. H. Gapopening transition in Dirac semimetal ZrTe_{5}. Phys. Rev. Lett. 126, 236401 (2021).
Liang, T. et al. Anomalous Hall effect in ZrTe_{5}. Nat. Phy. 14, 451–455 (2018).
Sun, Z. et al. Large zeeman splitting induced anomalous Hall effect in ZrTe_{5}. npj Quant. Mater. 5, 36 (2020).
Liu, Y. et al. Induced anomalous Hall effect of massive Dirac fermions in ZrTe_{5} and HfTe_{5} thin flakes. Phys. Rev. B 103, L201110 (2021).
Mutch, J. et al. Abrupt switching of the anomalous Hall effect by fieldrotation in nonmagnetic ZrTe_{5}. Preprint at https://arxiv.org/abs/2101.02681 (2021).
Lozano, P. M. et al. Anomalous Hall effect at the lifshitz transition in ZrTe_{5}. Preprint at https://arxiv.org/abs/2112.15227 (2021).
Zhang, Y. et al. Electronic evidence of temperatureinduced lifshitz transition and topological nature in ZrTe_{5}. Nat. Commun. 8, 15512 (2017).
Kamm, G. N., Gillespie, D. J., Ehrlich, A. C., Wieting, T. J. & Levy, F. Fermi surface, effective masses, and dingle temperatures of ZrTe_{5} as derived from the shubnikov–de haas effect. Phys. Rev. B 31, 7617–7623 (1985).
Izumi, M. et al. Shubnikovde haas oscillations and fermi surfaces in transitionmetal pentatellurides ZrTe5 and HfTe5. J. Phys. C: Solid State Phys. 20, 3691–3705 (1987).
Zhu, Z., Fauqué, B., Fuseya, Y. & Behnia, K. Angleresolved landau spectrum of electrons and holes in bismuth. Phys. Rev. B 84, 115137 (2011).
Wang, J. et al. Vanishing quantum oscillations in Dirac semimetal ZrTe_{5}. Proc. Natl. Acad. Sci. 115, 9145–9150 (2018).
Liang, T. et al. Evidence for massive bulk Dirac fermions in Pb_{1} − xSn_{x}Se from Nernst and thermopower experiments. Nat. Commun. 4, 2696 (2013).
Shahi, P. et al. Bipolar conduction as the possible origin of the electronic transition in pentatellurides: Metallic vs semiconducting behavior. Phys. Rev. X 8, 021055 (2018).
Salzmann, B. et al. Nature of native atomic defects in ZrTe_{5} and their impact on the lowenergy electronic structure. Phys. Rev. Materials 4, 114201 (2020).
Mott, N. F. & Davis, E. Electronic Processes in NonCrystalline Materials (Oxford, Clarendon Press, 1971).
Mott, N. F. Metalinsulator transitions (Taylor & Francis, London, 1990), 2nd edition edn.
Shklovskii, B. I. & Efros, A. L. Electronic properties of doped semiconductors (SpringerVerlag, New York, 1984).
Martino, E. et al. Twodimensional conical dispersion in ZrTe_{5} evidenced by optical spectroscopy. Phys. Rev. Lett. 122, 217402 (2019).
Dyakonov, M., Efros, A. & Mitchell, D. Magnetic freezeout of electrons in extrinsic semiconductors. Phys. Rev. 180, 813–818 (1969).
Chazalviel, J. N. & Solomon, I. Experimental evidence of the anomalous Hall effect in a nonmagnetic semiconductor. Phys. Rev. Lett. 29, 1676–1679 (1972).
Nozières, P. & Lewiner, C. A simple theory of the anomalous Hall effect in semiconductors. J. Phys. France 34, 901–915 (1973).
Biernat, H. & Kriechbaum, M. Anomalous Hall effect of n InSb at high magnetic fields. Phys. Status Solidi (B): Basic Res. 78, 653–657 (1976).
Behnia, K. Fundamentals of thermoelectricity (Oxford University Press, Oxford, 2015).
Skinner, B. & Fu, L. Large, nonsaturating thermopower in a quantizing magnetic field. Sci. Adv. 4, 2621 (2018).
Fu, B., Wang, H.W. & Shen, S.Q. Dirac polarons and resistivity anomaly in ZrTe_{5} and HfTe_{5}. Phys. Rev. Lett. 125, 256601 (2020).
Wang, C. Thermodynamically induced transport anomaly in dilute metals ZrTe_{5} and HfTe_{5}. Phys. Rev. Lett. 126, 126601 (2021).
Mutch, J. et al. Evidence for a straintuned topological phase transition in ZrTe_{5}. Sci. Adv. 5, eaav9771 (2019).
Zheng, G. et al. Transport evidence for the threedimensional Dirac semimetal phase in ZrTe_{5}. Phys. Rev. B 93, 115414 (2016).
Zhang, J. L. et al. Anomalous thermoelectric effects of ZrTe_{5} in and beyond the quantum limit. Phys. Rev. Lett. 123, 196602 (2019).
Acknowledgements
We thank K. Behnia, JH Chu, A. Jaoui, B. Skinner and B. Yan for useful discussions. We acknowledge the support of the LNCMICNRS, member of the European Magnetic Field Laboratory (EMFL). This work was supported by JEIPCollège de France, by the Agence Nationale de la Recherche (ANR18CE92002001; ANR19CE30001404), by a grant attributed by the Ile de France regional council and from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 636744). A.A. acknowledges funding from the Swiss National Science Foundation through project PP00P2_170544. The work at Brookhaven National Laboratory was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering, under Contract No. DESC0012704.
Author information
Authors and Affiliations
Contributions
B.F. and A.G. conducted the electrical, thermoelectrical and thermal conductivity measurements up to B = 17T. Highfield measurements have been conducted by M.L., M.M., D.V. and C.P. at LNCMIToulouse. Optical measurements have been conducted by A.A. and C.C.H. and analyzed by R.L. and A.A. Samples have been grown by Q.L. and G.G. Electrical contacts on the samples have been prepared by J.L.S. and B.F. B.F wrote the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Gourgout, A., Leroux, M., Smirr, JL. et al. Magnetic freezeout and anomalous Hall effect in ZrTe_{5}. npj Quantum Mater. 7, 71 (2022). https://doi.org/10.1038/s4153502200478y
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s4153502200478y
This article is cited by

Highfield immiscibility of electrons belonging to adjacent twinned bismuth crystals
npj Quantum Materials (2024)

Signatures of a magneticfieldinduced Lifshitz transition in the ultraquantum limit of the topological semimetal ZrTe5
Nature Communications (2022)