betaFeedback

WorksPhysics and Astronomy

Quantum oscillations revealing topological band in kagome metal ScV6Sn6

Abstract

Compounds with kagome lattice structure are known to exhibit Dirac cones, flat bands, and van Hove singularities, which host numerous versatile quantum phenomena. Inspired by these intriguing properties, we investigate the temperature and magnetic field dependent electrical transports along with the theoretical calculations of ScV6Sn6, a nonmagnetic charge density wave (CDW) compound. At low temperatures, the compound exhibits Shubnikov-de Haas quantum oscillations, which help to design the Fermi surface (FS) topology. This analysis reveals the existence of several small FSs in the Brillouin zone, combined with a large FS. Among them, the FS possessing Dirac band is a non-trivial and generates a non-zero Berry phase. In addition, the compound also shows the anomalous Hall-like behaviour up to the CDW with the CDW phase, ScV6Sn6 presents a unique material example of the versatile HfFe6Ge6 family and provides various promising opportunities to explore the series further.

The kagome lattice gives rise to unavoidable exotic topological electronic states namely Dirac point [1, 2], Van Hove singularity [3, 4] and flat band [1, 2]. These features have been extensively studied by different spectroscopy experiments in various kagome compounds. Depending on band filling and interactions, they accommodate a number of unconventional quantum phases, topological band structure [5, 6, 7], Chern insulator [8], unconventional superconductivity [9, 10] and charge density wave (CDW) [11, 12, 13] [14]. The profound impacts of the kagome lattice have recently been highlighted by the discovery of CDW below 90 K in nonmagnetic AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} and antiferromagnetic hexagonal-FeGe [11]. The charge orders in these systems are distinct. In AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} [14–18], it is chiral, whereas in FeGe [11] [19] magnetism is mediated. Similar to AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} and FeGe, the CDW transition temperature (at 92 K) was recently discovered in the ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} compound [20]. ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} is the only known compound from the vast hexagonal HfFe6Ge6\mathrm{HfFe}_{6}\mathrm{Ge}_{6} family to exhibit a CDW phase transition, which is a first-order-like transition with propagation vector (1/3,1/3,1/3)(1/3, 1/3, 1/3) revealed by early X-ray and neutron experiments [20]. The strongly coupled out-of-plane lattice dynamics suggests an unconventional nature of the CDW phase, which is different from the in-plane lattice dynamics in AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} [12]. Furthermore, CDW of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} is easy to tune [21] and exhibits various microscopic features such as critical role of phonons [22–25], large spin Berry curvature [26], partial bandgap opening [26–28], hidden magnetism [29]. As the first such member of the versatile kagome family of HfFe6Ge6\mathrm{HfFe}_{6}\mathrm{Ge}_{6}-type compounds, it is worthwhile to study the temperature and magnetic field dependent electrical transport properties of high-quality single crystals of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} in order to obtain information about the Fermi surface (FS) and lattice dynamics, reporting in this letter.

ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} crystallizes at room temperature in a hexagonal centrosymmetric structure with a P6/mmmP6/mmm space group. The V atoms form a kagome lattice in the abab-plane (Figure 1(a) left), while the Sn atoms sit above and below, separating the lattice [20]. High quality hexagonal single crystals (Figure 1(b), inset) were grown by the flux method, see supplemental information (SI) [30].

Kagome lattice, Laue pattern, I-V characteristic, and CDW phase transition of ScV6Sn6

Figure 1. Kagome lattice, Laue pattern, I-V characteristic, CDW phase transition of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6}. (a) Kagome lattice of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} viewed along the cc-axis (left) and recorded single crystal Laue patterns along the cc-axis. (b) I-V characteristics at various fields at 2K. The inset is an optical image of the crystal with Cartesian coordinate. (c) Magnetic susceptibility along x^\hat{x}. Resistivity in zero-field (d) along y^\hat{y} and (c) along z^\hat{z}, where the sudden jumps show CDW phase transition.

To facilitate further measurements, the crystal orientations were marked in Cartesian coordinates x^\hat{x}, y^\hat{y}, z^\hat{z}, which correspond to (21ˉ1ˉ0)(2\bar{1}\bar{1}0), (011ˉ0)(01\bar{1}0), (0001)(0001) crystallographic directions in the hexagonal crystal structure, respectively. Laue X-ray diffraction patterns were recorded along z^\hat{z} and show a clear six-fold symmetry (Figure 1(a), right). The patterns fit well with the lattice parameters a=b=5.4733(5)a=b=5.4733(5) Å and c=9.1724(8)c=9.1724(8) Å derived from the fitting of powder X-ray patterns (Fig. S1) [30,31]. The measured chemical compositions from the energy dispersive X-ray spectrum are close to the stoichiometric atomic ratio of Sc:V:Sn=1:5.97:6.07\mathrm{Sc}:\mathrm{V}:\mathrm{Sn}=1:5.97:6.07.

We further cut the crystals in defined orientations of Hall-bar geometry and fabricated fourprobe Pt-wire contacts with highly conductive Ag paint. The II-VV characteristics at 2 K in different fields are linear (Figure 1(b)), confirming the high quality of the ohmic contacts. The appearance of CDW in this compound is a key property, and we found it in the temperature-dependent magnetization and the electrical resistivity (Figure 1(c-Figure 1e)). A sharp transition appears at 92 K, depicting the CDW transition temperature similar to the previous reports [19, 20]. The measured magnetic susceptibility in Figure 1(c) describes a weak Pauli paramagnetic-like behavior, which is different from its structural sister compound YV6_6Sn6_6 [32]. The zero-field electrical resistivity shows a metallic behavior with temperature, and both ρyy\rho_{yy} and ρzz\rho_{zz} drop suddenly at the CDW phase transition due to the large softening of the acoustic phonon modes (Figure 1(d, Figure 1e)) [23,24]. The typical values of ρyy\rho_{yy} (ρzz\rho_{zz}) at 2 K are found to be 1×10−51 \times10^{-5} (4.58×10−54.58 \times10^{-5}) Ω\Omega cm and the resulting residual resistivity ratio (RRR=ρ300K/ρ2KRRR = \rho_{300\mathrm{K}}/\rho_{2\mathrm{K}}) is 8.9(4.3). Noticeably, the ρzz\rho_{zz} changes significantly than the ρyy\rho_{yy} at the CDW phase transition.

After the basic characterization of the crystals, we now focused on the magnetic field, BB-dependent transverse resistivity, which were measured at different temperatures and angles in the field of ±9\pm9 T. The angular dependence of the field is defined as θ\theta (=x^→z^= \hat{x} \to\hat{z}) and ϕ\phi (=x^→y^= \hat{x} \to\hat{y}), where 0∘=B∥x^0^\circ= B \parallel\hat{x}, 90∘=B∥y^90^\circ= B \parallel\hat{y} or z^\hat{z}. First, we describe the behavior of ρzz\rho_{zz} when B∥x^B \parallel\hat{x} (θ\theta and ϕ=0\phi= 0), and the measured data are shown in Figure S6 [30]. At low temperatures, ρzz\rho_{zz} exhibits clear Shubnikov de-Haas (SdH) quantum oscillations which are a striking feature that helps to probe the low-energy bands. At 2 K, the quantum oscillation starts from the field around 1.5 T, as seen in Figure 2(a) for the dρzz/dBd\rho_{zz}/dB plot and it is easily visible up to 12 K (Figure S6 [30]). The magnetoresistance MR (=ρzz(B)/ρzz(0)−1=\rho_{zz}(B)/\rho_{zz}(0) - 1) is estimated, and the highest value is found to be 170 % at the lowest temperature of 2 K (Figure S6) [30]. Unlike the other typical semimetals, e.g., NbP, the present compound ScV6_6Sn6_6 does not show an extremely large MR [33]. However, the SdH oscillation indicates a low effective mass of the electron charge carrier in the CDW phase [23,24]. To obtain the amplitude of the SdH oscillations, a smooth polynomial background was subtracted from the measured ρzz\rho_{zz}. The resultant Δρzz\Delta\rho_{zz} at several temperatures is plotted as a function of 1/B1/B, as shown in Figure 2(b). As expected, the oscillations are periodic in 1/B1/B and they arise from quantization of the energy level further, forming Landau levels (LLs) [34]. The analysis of the temperature and field-dependent periodic oscillations provides insight into the fermiology of the compound and the associated physical characteristics of the charge carriers.

Shubnikov–de Haas oscillations and Berry phase

Figure 2. Shubnikov–de Haas (SdH) oscillations and Berry phase. (a) Longitudinal ρzz\rho_{zz} at 2 K and its first derivative with field showing quantum oscillations. (b) Background subtracted SdH oscillations amplitude at several temperatures. (c) Corresponding FFT amplitude exhibiting α\alpha, β\beta and γ\gamma frequencies. (d) FFT amplitude fit estimating effective mass. (e) SdH oscillations when B∥z^B \parallel\hat{z} and their FFT spectrum (inset) showing only the γ\gamma frequency in the field window of 5-8 T. (f) Intercept of the Landau fan diagram revealing non-zero Berry phase. (g) Angular dependent of the Berry phase.

Longitudinal magnetoresistance and resistivity

Figure S6. Longitudinal Magnetoresistance (MR) and resistivity. (a) MR for different temperatures for I∥z^I \parallel\hat{z}, B∥x^B \parallel\hat{x}. (b) first derivative from magnetic field for ρzz\rho_{zz} below 20 K for clarify the SdH oscillations. (c) Fast Fourier transformation (FFT) results of the observed SdH quantum oscillations of Δρzz\Delta\rho_{zz}. Frequency range for reconstructing each oscillation are plotted in green (α\alpha), orange (β\beta), and azure (γ\gamma) colors. (d) Reconstructed SdH oscillations for α\alpha-, β\beta- and γ\gamma-pocket by bandwidth pass processes at 2 K.

In the fast Fourier transform (FFT) of the SdH oscillations presented in Figure 2(c), we detected the existence of three distinct frequencies: Fα=10F_{\alpha}=10 T, Fβ=28F_{\beta}=28 T and Fγ=50F_{\gamma}=50 T. These frequencies indicate the presence of small FSs. Compared to AV3_3Sb5_5 (A = K, Rb, Cs), the splitting of FS in ScV6_6Sn6_6 is simpler. At least 9 FSs have been detected in CsV3_3Sb5_5, with their belonging frequency ranging from 10 T to 2000 T [35–38]. Lifshitz–Kosevich (L-K) theory is used extensively to analyze the physical properties of FSs [34], and it fully explains the SdH oscillations. The effective mass, Dingle temperature and phase factor are evaluated using L-K equation, where Δρ∝RTRDcos⁡[2π(FB+γ−δ)]\Delta\rho\propto R_T R_D \cos\left[2\pi\left(\frac{F}{B}+\gamma-\delta\right)\right]. Specifically, RT=14.69m∗T/Bsinh⁡(14.69m∗T/B)R_T=\frac{14.69m^{*}T/B}{\sinh(14.69m^{*}T/B)} and RD=exp⁡(−14.69m∗TD/B)R_D=\exp(-14.69m^{*}T_D/B) are the cyclotron effective mass term and Dingle term, respectively; in which m∗m^{*} (m0m_0) is the effective (bare) mass of the electron, BB is the average field used in the FFT, TDT_D is the Dingle temperature. The phase factor γ−δ\gamma-\delta, which is directly related to the Berry phase within δ\delta, equals to 0 for the 2D system and ±18\pm\frac{1}{8} for the 3D system (±\pm corresponds to the contribution from the minimal/maximal cross section) [36,39,40]. Among the various parameters of FS that can be estimated using L-K formula, the estimation of the Berry phase is very promising for proving the non-trivial topology of the band. Accordingly, we designed the Landau fan diagram by assigning LLs to the oscillatory extrema. The integer nn-th LLs are assigned to maxima when ρxx≫ρyx\rho_{xx}\gg\rho_{yx}, while they are assigned to minima when ρxx≪ρyx\rho_{xx}\ll\rho_{yx} [36,39,41]. For ScV6_6Sn6_6, where ρzz∼10ρyz\rho_{zz}\sim10\rho_{yz} which satisfies the condition of ρxx≫ρyx\rho_{xx}\gg\rho_{yx}, the integer nn-th are assigned to the maxima and the half integer (n+12)(n+\frac{1}{2})-th are assigned minimum of Δρzz\Delta\rho_{zz}, as shown in Figure 2(e, Figure 2f). The Landau indices nn and 1/B1/B satisfy the Lifshitz-Onsager relationship, which is described by the equation n=FB+γ−δn=\frac{F}{B}+\gamma-\delta. The slope of the plot between nn and 1/B1/B provides the oscillatory frequency, while the intercept offers the information on the Berry phase. To attend an accurate estimation of Berry phase of FγF_\gamma, we selected a field window between 5 to 8 T to exclude the weak FαF_\alpha oscillatory component (Figure 2(e) inset). From Figure 2(f), the intercept of the Landau fan diagram for FγF_\gamma in terms of the phase factor γ−δ\gamma-\delta is estimated to be -0.101, indicating a nontrivial Berry phase [42,43]. Similarly, we obtained the Berry phase further for the angular dependence of the field from x^\hat{x} to z^\hat{z} without interference from the other exiting frequencies, and the values are given in Figure 2(g). The figure indicates the Berry phase factor γ−δ\gamma-\delta falls within the range of 0±180 \pm\frac{1}{8} for θ≥60∘\theta\ge60^\circ. While it is possible to estimate the Berry phase factor θ>20∘\theta> 20^\circ, since the FβF_{\beta} is absent, but the SdH amplitude declines rapidly. At all other angles of θ\theta and ϕ\phi, the FβF_{\beta} is consistently present as can be seen in Figure 4(b), which intervenes the amplitude in terms of the beating patterns [44,45], as can be seen in Figure S9 and (Figure S9) [30]. One method to separate the amplitude corresponding each frequency is to use the band-passing filter as shown in Figure S6 and (Figure S6) [30,46], but may not be very reliable. We also tried to estimate to Berry phase related to FαF_{\alpha} and FβF_{\beta}. As we can see, FαF_{\alpha} shows very weak oscillatory amplitude with limited cycle up to 9 T. One the other hand, the presence of the beating patterns harms the amplitude of the FβF_{\beta}, which only appears for θ≤20∘\theta\le20^\circ and after that the FβF_{\beta} vanishes. These together make difficult to carry out the Berry phase analysis for FαF_{\alpha} and FβF_{\beta}.

SdH oscillations and fermiology

Figure 4. SdH oscillations and fermiology. (a) SdH oscillation frequency at different rotating angles θ\theta (=x^→z^=\hat{x}\to\hat{z}) and ϕ\phi (=x^→y^=\hat{x}\to\hat{y}), where dotted lines represent their tracking with angle. The insets show a sketch of the field rotation. (b) Angular dependence of Fα,β,γF_{\alpha,\beta,\gamma}, where Fα,βF_{\alpha,\beta} do not change while FγF_{\gamma} shifts slightly. Solid lines are calculated frequencies corresponding. The error bar is taken from the half width of the half maximum value of the frequency peaks. (c) Brillouin zone and three-dimensional (3D) Fermi surfaces (FSs) of the bands at the Fermi energy. (d-g), Enlarged FSs, including the FS at MM point (d, yellow), the banana-shaped α\alpha-pocket along MM-LL (e, red), the spindle-shaped inner β\beta-pocket at KK point (f, yellow), and the apple-shaped outer γ\gamma-pocket at KK point (g, blue).

Resistivity and SdH oscillations

Figure S9. Resistivity, SdH oscillations. (a) and (b) are resistivity when magnetic field is rotating from x^\hat{x} to z^\hat{z} and x^\hat{x} to y^\hat{y}, respectively. The inset of (b) is the enlarged view for clarity. (c) and (d), background subtracted resistivity Δρzz\Delta\rho_{zz} as functions of 1/B1/B for the both configurations mentioned in (a) and (b) respectively, showing the periodic oscillations.

To determine the effective mass for the carriers corresponding to each FS, the temperature dependence of the amplitude of each peak in the FFT is plotted, and then it is fitted with RTR_{\mathrm{T}} term (Figure 2(d)) from the L-K formula. From these fittings, we find the values of mα,β,γ∗=0.105 m0, 0.128 m0, 0.184 m0m^{*}_{\alpha,\beta,\gamma} = 0.105\ m_{0},\ 0.128\ m_{0},\ 0.184\ m_{0}. These masses are very low (even lower than CsV3_{3}Sb5_{5} [35,36]). Furthermore, the Dingle temperature TDT_{\mathrm{D}} can be estimated from the semi-log plot described by ln⁡(A/RT)∝−14.69m∗TD/B\ln(A/R_{\mathrm{T}}) \propto-14.69m^{*}T_{\mathrm{D}}/B (Figure S8) [30], where AA corresponds to the amplitude of the SdH oscillations. We found TDT_{\mathrm{D}} to be 2.5 K for γ\gamma-pocket and the corresponding quantum scattering time τq\tau_{\mathrm{q}} is ∼5×10−13\sim5 \times10^{-13} s. For the Fα,β,γ=10, 28, and 50F_{\alpha,\beta,\gamma} = 10,\ 28,\ \text{and}\ 50 T, the equivalent periodicities 1/(Fα,β,γ)=0.1, 0.036, and 0.02 T−11/(F_{\alpha,\beta,\gamma}) = 0.1,\ 0.036,\ \text{and}\ 0.02\ \mathrm{T}^{-1}, which are further related to the cross-sectional area AFA_{\mathrm{F}} of FS by the Onsager relation F=(Φ0/2π2)AFF = (\Phi_{0}/2\pi^{2})A_{\mathrm{F}}, where the value of magnetic flux quantum Φ0=2.068×10−15\Phi_{0} = 2.068 \times10^{-15} Wb. The corresponding area AFα,β,γ=0.001, 0.0027, and 0.0048 A˚−2A_{\mathrm{F}_{\alpha,\beta,\gamma}} = 0.001,\ 0.0027,\ \text{and}\ 0.0048\ \text{\AA}^{-2}. Assuming a circular cross-section of these FSs for simplicity, the very small Fermi vectors kFα,β,γ=0.017 A˚−1, 0.029 A˚−1, 0.039 A˚−1k_{\mathrm{F}_{\alpha,\beta,\gamma}} = 0.017\ \text{\AA}^{-1},\ 0.029\ \text{\AA}^{-1},\ 0.039\ \text{\AA}^{-1} are found; and their corresponding Fermi velocities vFα,β,γv_{F\alpha,\beta,\gamma} are 1.92×1051.92 \times10^5 m s−1^{-1}, 2.64×1052.64 \times10^5 m s−1^{-1}, 2.45×1052.45 \times10^5 m s−1^{-1}. Additional relevant experimental parameters of the FSs are summarized in Table S1 [30].

Dingle plot of $F_{\gamma}$ when $B \parallel \hat{x}$

Figure S8. Dingle plot of FγF_{\gamma} when B∥x^B \parallel\hat{x}.

FF (T)m∗m^* (m0m_0)AFA_{\mathrm{F}} (Å−2^{-2})kFk_{\mathrm{F}} (Å−1^{-1})vFv_{\mathrm{F}} (10510^5 ms−1^{-1})ll (nm)TDT_{\mathrm{D}} (K)τq\tau_{\mathrm{q}} (10−1310^{-13} s)μq\mu_{\mathrm{q}} (10310^3 cm2^2V−1^{-1}s−1^{-1})
FαF_{\alpha} : 100.1050.0010.0171.92---
FβF_{\beta} : 280.1280.00270.0292.64----
FγF_{\gamma} : 500.1840.00480.0392.451232.435.004.78

Table S1. Fermi surface parameters (effective mass m∗m^*, Fermi vector kFk_{\mathrm{F}}, Fermi velocity vFv_{\mathrm{F}}, scattering length ll, Dingle temperature TDT_{\mathrm{D}}, quantum scattering time τq\tau_{\mathrm{q}}, quantum mobility μq\mu_{\mathrm{q}}) extracted from the SdH quantum oscillations from Figure 3 and Figure S6

To capture the topology in electronic band structure of ScV6Sn6\mathrm{ScV}_6\mathrm{Sn}_6, we performed detailed band structure calculations by using density functional theory (DFT). Due to the presence of CDW transition, we transformed the band structure of low temperature (LT) phase into the high-temperature (HT) phase, and the LT phase (CDW phase) has been characterized at a temperature of 50 K and the lattice parameters are taken from the previous report [20]. From Figure S11 [30], the unfolded band structures of LT and HT phases are in a good agreement with the angle-resolved photoemission spectrum (ARPES) experiment [27]. Among the two phases, the major change in the electronic band structure appears only along the AA-LL-HH direction (where the band opens a gap because of the CDW transition [27]), while the Van Hove singularity bands at the MM point and the Dirac bands at the KK point do not change. From a closer view, a fascinating feature depicting in Figure 3(a) is the band folding phenomenon, whereby the original KK point at coordinates (1/3,1/3,0)(1/3, 1/3, 0) transmutes into a folded K1K1 point situated at (1/3,1/3,−2/3)(1/3, 1/3, -2/3). This figure unambiguously reveals that the K1K1 point intrinsically possesses a substantial band gap near the Fermi energy (EFE_{\mathrm{F}}). Furthermore, both the conduction and valence bands are completely split at this point, implying the absence of topologically protected Dirac points. Utilizing the transformation matrix, we ascertain that the original MM point at (1/2,0,0)(1/2, 0, 0) and the Γ\Gamma point at (0,0,0)(0, 0, 0) have been mapped to the new M1M1 point at (1/2,0,−1/2)(1/2, 0, -1/2) and Γ\Gamma point at (0,0,0)(0, 0, 0), respectively. As a result, we discovered that the genuine Dirac point (labeled as DP in Figure 3b) has been mapped to the position of (0.341666,0.316666,−0.658327)(0.341666, 0.316666, -0.658327). From this magnified view of the band structure near this point, its Dirac nature is clearly confirmed. Similar to the Dirac points in graphene systems, this point develops a band gap when considering spin-orbit coupling as shown in Figure 3(c), and gives rise to non-trivial topological properties. Therefore, we believe that the non-trivial π\pi Berry phase extracted in the SdH oscillation actually comes from the electron pockets formed in the energy bands where the DP is located.

Band structure in charge density wave phase

Figure 3. Band structure in charge density wave (CDW) phase. (a) Band structure of ScV6_6Sn6_6 CDW phase with the presence of spin orbital coupling (SOC). The structural parameters are the same with the report [20] at 50 K. The magnified view of the band structure around K1 locating the Dirac point (DP) (b) without the SOC and (c) with the SOC.

Electronic band structures

Figure S11. Electronic band structures. (c) Band structure of ScV6Sn6ScV_{6}Sn_{6} based on the crystal structure in high-temperature (HT) phase (black) and low-temperature (LT) CDW phase (blue) when spin-orbit coupling (SOC) is ignored. (b) Band structures with SOC. In a and b, the band structures of LT phase are unfolded to match the band structure of HT phase for comparison. (c, d) Orbitals projected band structure for HT phase with and without SOC, respectively.

To know more information about the topography of the FS, we measured the angular θ(=x^→z^)\theta(=\hat{x}\to\hat{z}) and ϕ(=x^→y^)\phi(=\hat{x}\to\hat{y}) dependent oscillations. The rotating angles of field with the crystal axes are sketched in the insets of Figure 4(a), where 0∘=B∥x^0^\circ=B\parallel\hat{x}, 90∘=B∥y^90^\circ=B\parallel\hat{y} or z^\hat{z}. The angular dependent of FFT is plotted in Figure 4(a). As we mentioned earlier, the three frequencies namely α\alpha, β\beta and γ\gamma appear when B∥x^B\parallel\hat{x}. Depending upon the shape of FS, the movement of frequency depends on the rotating angle. From Figure 4(a), we carefully tracked the position of all visible frequencies with the rotating angles and demonstrated in Figure 4(b). This indicates that when the field BB moves either from the x^→y^\hat{x}\to\hat{y} or x^→z^\hat{x}\to\hat{z}, the frequencies α\alpha, β\beta and γ\gamma change slightly. Noticeably, when the θ>20∘\theta>20^\circ, the β\beta disappears. This helps further to improve the accuracy of Berry phase analysis related to γ\gamma. In order to allocate the experimentally observed frequencies, we theoretically calculated 3D FSs and their corresponding frequency from the unfolded band structure of the LT phase. As we can see in Figure 4(b), all observed frequencies are accurately replicated by the calculated 3D FSs, except for higher frequency (green line). In addition, the calculated FβF_\beta is present at all angles, but it disappears θ>20∘\theta>20^\circ in the experiment, which needs a further investigation. Figure 4(c) shows 3D FSs within the BZ, in which a large electron pocket (blue color) around Γ\Gamma covers the whole BZ together with a small banana-shaped FS (red color) along the MM-LL direction, a small apple-shaped FS (blue color) at KK, and a larger FS at MM (yellow color). Another spindle-shaped FS at KK which comes from the same band with the FS at MM and is covered by the apple-shaped pocket. After comparing with DFT results, we found that the FαF_{\alpha} belongs to the banana-shaped FS (Figure 4(e)). The FβF_{\beta} belongs to the spindle-shaped FS (Figure 4(f)). The FγF_{\gamma} is part of the apple-shaped FS that is contributed by the Dirac bands (Figure 4(g)), exhibiting the non-trivial Berry phase. The calculated high frequency (green solid line in Figure 4(b)) comes from the FS (Figure 4(d)) at MM point, which is not detectable up to 9 T in the present measurements.

After exploring the striking features of the FS, we now get the insight of classical Drude transport from the Hall resistivity ρzy\rho_{zy} for current I∥y^I \parallel\hat{y} and B∥x^B \parallel\hat{x}. From these data, the Hall constant RHR_{\mathrm{H}} (=ρzy/B=\rho_{zy}/B) was calculated by considering the high field slope between 6–9 T at different temperatures (Figure 5(a), inset), resulting in electron density nn (=1/eRH=1/eR_{\mathrm{H}}) and mobility μ\mu (=RH/ρyy=R_{\mathrm{H}/\rho_{yy}}) of 2.4×10212.4 \times10^{21} cm−3^{-3} and 265 cm2^{2} V−1^{-1} s−1^{-1} at 2 K, respectively (Figure S3) [30]. The Drude scattering time, τtr=μm∗/e=2.1×10−14\tau_{\mathrm{tr}}=\mu m^{*}/e=2.1 \times10^{-14} s, where m∗m^{*} is 0.139 m00.139\,m_{0} (the average of m∗m^{*}) and ee is the elementary charge. The ratio of τc/τtr\tau_{\mathrm{c}}/\tau_{\mathrm{tr}} (∼25\sim25) is large, where τc\tau_{\mathrm{c}} and τtr\tau_{\mathrm{tr}} are related to electron-electron and electron-phonon scattering processes, respectively. Besides the usual high and low angle scattering mechanisms, the large lowering value τtr\tau_{\mathrm{tr}} could be related to the new phonon modes, arising from promoting the electron-phonon coupling crucially in the CDW phase [22–24].

Anomalous Hall-like behavior in ScV6Sn6, with panels showing field-dependent Hall resistivity, temperature- and angular-dependent anomalous Hall conductivity, and proposed microscopic sources

Figure 5. Anomalous Hall-like behavior in ScV6_6Sn6_6. (a) Field-dependent measured anomalous behavior of Hall resistivity, ρzy\rho_{zy} at various temperatures (inset) and extracted Hall resistivity ρzyA\rho_{zy}^{A}. (b) Temperature and (c) angular dependent corresponding anomalous Hall like conductivity. Electrons picking up anomalous velocity from the different sources and their typical example. (d) Non-zero Berry phase from ferromagnetic spins. (e) Topological orbital moment due to spins chirality. (f) Formation of local loop currents.

Resistivity, MR, carrier density and mobility, Hall resistivity

Figure S3. Resistivity, MR, carrier density and mobility, Hall resistivity. (a) Temperature-dependent resistivity ρyy\rho_{yy} at various fields when I∥y^,B∥x^I \parallel\hat{y}, B \parallel\hat{x}. (b) Magnetoresistance as functions of magnetic field at various temperatures. (c) Estimated carrier concentration and mobility as functions of temperature. The sudden change is related to CDW phase transition, which is also highlighted by the vertical black dashed line. (d) Hall resistivity as functions of magnetic field at different rotating angles. The field is rotated from the x^\hat{x} to the y^\hat{y} directions as shown in the inset.

The field dependent ρzy\rho_{zy} shows a non-linear behavior, which is primarily thought to be the presence of multi-carriers in a nonmagnetic system such as ScV6_{6}Sn6_{6}, since the multiple bands are around the EFE_{\mathrm{F}}. To clarify this apparent reason, we fitted both the Hall and longitudinal conductivities by using the standard two-band model (see Figure S5 in SI) [30,47–50]. Both fits are seemingly good, but their fitted parameters are very different from each other. The fitting of the σyz\sigma_{yz} reflects both electron carriers, while the fitting of σyy\sigma_{yy} shows both electron and hole carriers. For the implementation of a reliable and accurate model, the parameters from the both fittings must be the same within the error bar as reported in the typical multiband systems [49,50]. Therefore, given the type of carriers, the concentrations and mobilities are highly inconsistent, indicating that the two-band model may not be applicable to the nonlinear Hall effect for ScV6_{6}Sn6_{6}. It is also worthy to note here that the Hall resistivity from a typical multi-band system is nonlinear close to zero field, while it is become linear at high field [33,51–53]. In contrast, the present compound ScV6_{6}Sn6_{6} shows the linear Hall close to zero field and it bends at high field (>2>2 T) like a typical soft ferromagnetic system [54,55].

Two-band model fitting

Figure S5. Two-band model fitting. Individual fitting of (a) Hall conductivity by Eq. (2) and (b) longitudinal conductivity by Eq. (3) at T=10T = 10 K, respectively. (c-d) Simultaneously fitting of Hall and longitudinal conductivities.

By considering the anomalous behavior naively, we subtracted the linear part from the high field and the anomalous part of the Hall resistivity at different temperatures is shown in Figure 5(a). Like in soft ferromagnet, an anomaly in the ρzy\rho_{zy} (a sharp increase at a lower field followed by saturation with a further increase in the field) is observed, which is attributed to anomalous Hall like behavior. Such an anomaly usually appears as a hallmark in magnetic systems and it resembles the magnetization curve, wherein magnetic spins play the role (Figure 5(d, Figure 5e)). However, magnetization is linear with the field (see Figure S2 in SI) [30] and does not follow the Hall resistivity in the present case. The as calculated value of anomalous Hall conductivity (AHC ≈ρzyAρzzρyy\approx\frac{\rho^{A}_{zy}}{\rho_{zz}\rho_{yy}} since ρyy≠ρzz\rho_{yy} \ne\rho_{zz}) at 2 K is 3.2×103 Ω−1 cm−13.2 \times10^{3}\ \Omega^{-1}\,\mathrm{cm}^{-1} which is smaller than AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} systems [14,15,38]. The AHC remains almost constant up to 45∘45^\circ and starts to decrease with increasing angle further (Figure 5(c)), while it sustains up to the CDW transition temperature. In contrast to the involvement of magnetism (Figure 5(d), (Figure 5e)), one of the possible explanations could be similar to the origin in AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} (Figure 5(f)), where the CDW forms a current loop and breaks the time reversal symmetry (TRS) [16,56,57], since ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} also breaks the TRS measured by the muon spectroscopy [29] and to speculate such loop current is a bit premature. In the present scenario, it is difficult to understand the observed Hall behavior of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6}, and needs to do further study. Moreover, the behavior of ρyz\rho_{yz} is closer to ZrTe5\mathrm{ZrTe}_{5}, assuming the presence of a single Dirac band that splits into a pair of Weyl points in presence of the magnetic field [58]. However, the several bands are present in ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6}, where the concept of Weyl points is not easily applicable.

Magnetic measurements

Figure S2. Magnetic measurements. (a), and (c), Temperature-dependent magnetization at several magnetic fields B∥x^B \parallel\hat{x} and B∥z^B \parallel\hat{z}, respectively. (b), and (d), Isothermal magnetization as a function of magnetic fields B∥x^B \parallel\hat{x} and B∥z^B \parallel\hat{z}, respectively for different temperatures.

In conclusion, with the help of field-dependent electrical transports and theoretical calculations, we have revealed the fermiology of ScV6_6Sn6_6. Along with a large FS, the BZ is comprised of the several other small electron FSs, except for the one tiny hole FS. The small FSs exhibit SdH oscillations, resulting in the small effective mass, and the symmetric and asymmetric nature of the FSs for electron and hole, respectively. The electron Dirac band located at the KK point gives rise to the non-zero Berry phase, proving that it is a non-trivial topological FS. Furthermore, the non-linear Hall, which resembles the anomalous Hall, does not fit well with the two-carrier model. Finding the microscopic origins for the anomalous Hall-like signal remains an important issue that requires further theoretical and experimental investigations.

Note added. During the preparation of our work, a similar work including anomalous Hall-like behavior in ScV6Sn6 was found [59].

Acknowledgments

This work was financially supported by the Deutsche Forschungsgemeinschaft (DFG) under SFB1143 (project no. 247310070), the Würzburg-Dresden Cluster of Excellence on Complexity and Topology in Quantum Matter—ct.qmat (EXC 2147, project no. 390858490) and the QUAST-FOR5249-449872909. S.R. thanks the Alexander von Humboldt Foundation for a fellowship.

Magnetic and Electrical Measurements

Temperature-dependent susceptibility (2–300 K) under various magnetic fields was measured in zero field cooled and field cooled configuration in a Magnetic Properties Measurement System (MPMS, Quantum Design Inc.) equipped with a Superconducting Quantum Interference Device - Vibrating Sample Magnetometer (SQUID-VSM) option. Temperature-dependent four-probe transverse and Hall resistivities were measured on a standard rotating sample holder in Physical Property Measurement System (PPMS, Quantum Design Inc.). The alternative current transport (ACT) option was used for measurement with an excitation current of 1.5 mA at a fixed frequency of 93 Hz. Data were collected in the temperature range of 2–300 K, and the magnetic field range of -9 to 9 T in sweep mode.

Electronic Band Structure Calculations

Ab initio calculations were performed using density functional theory (DFT) implemented in the Vienna Ab initio Simulation Package (VASP) [60,61]. The projector augmented wave method [62] and the generalized gradient approximation (GGA) with Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional [63] were used for the calculation. Here, we used refined lattice parameters for the band structure calculations, including the pristine structure and the CDW structure. A 15×15×715 \times15 \times7 kk-mesh was used to sample the Brillouin zone of the pristine structure, with an energy cut-off of 500 eV. For the CDW structure, an 8×8×88 \times8 \times8 kk-mesh was used, with an energy cut-off of 450 eV. The CDW band structures were unfolded using the VASPKIT toolkit [64]. To obtain the quantum oscillation frequencies, we constructed a Wannier tight-binding Hamiltonian using the Wannier90 code [65], including Sc-3d, V-3d, and Sn-5p orbitals.

Supplementary information

Single crystal growth

High-quality single crystals of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} were grown by the flux method [1]. High-purity starting elements Sc (99.99 %, Alfa Aesar), V (99.98 %, Alfa Aesar), and Sn (99.999 %, Alfa Aesar) were placed in an alumina crucible in a molar ratio of 1 : 10 : 60 and then sealed in an evacuated quartz tube. The tube was then slowly heated to 1373 K, maintained for 10 h, and cooled down to 973 K over 400 h. Hexagonal shape crystals with silvery surfaces and a typical size of 2×2×1 mm32 \times2 \times1\ \mathrm{mm}^{3} were obtained after centrifugation. Hall bars were cut from the crystals and polished before measurements were taken. The crystal orientations were determined by using the Laue diffractometer equipped with the Mo- X-ray source.

Crystal structure characterization

As grown crystals of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} were grinded into the fine powders to perform the powder X-ray diffraction (PXRD) experiment. As shown in Figure S1, the observed diffraction pattern at room temperature can be well refined using the P6/mmmP6/mmm space group by the FullProf program [2]. After the refinement, the lattice parameters are a=b=5.47332(5)a=b=5.47332(5) Å and c=9.17248(8)c=9.17248(8) Å, which are very close to the lattice parameters of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} reported in the previous report [1]. Only a tiny impurity peak is observed due to the Sn flux as many pieces were grind together for PXRD.

Crystal structure and powder X-ray diffraction pattern

Figure S1. Crystal structure, powder x-ray diffraction pattern. The powder x-ray diffraction pattern of grinded ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} single crystals. The inset is a schematic picture of the crystal structure in 3D view. The red asterisk is marked for impurity, possibly Sn.

Electromagnetic measurements

Temperature-dependent susceptibility (2–300 K) under various magnetic fields was measured in zero field cooled (ZFC) and field cooled (FC) configuration in an MPMS SQUID-VSM (Quantum Design). Temperature-dependent four-probe transverse and Hall resistivities were measured on a standard rotating sample holder in PPMS (Quantum Design). The alternative current transport (ACT) option was used for measurement with an excitation current of 1.5 mA at a fixed frequency of 93 Hz. Data were collected in the temperature range of 2–300 K, and the magnetic field range of -9 to 9 T in sweep mode.

For the magnetic property measurements of ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6}, we selected flux-free crystals and a typical optical image as shown in the inset of Figure S2. For ease of further reference, we marked the crystal axes in the Cartesian coordinates x^\hat{x}, y^\hat{y}, z^\hat{z}, which are correspond to (21ˉ1ˉ0)(2\bar{1}\bar{1}0), (011ˉ0)(01\bar{1}0), (0001)(0001) crystallographic directions in the hexagonal crystal structure. The isomagnetic and isothermal magnetic properties were measured by applying a field along x^\hat{x} and z^\hat{z}, the measured data are shown in Figure S2 (Figure S2) and S2 (c) (d), respectively. The magnetization decreases linearly with temperature and then suddenly drops at 92 K. The sudden decrease marks the charge density wave (CDW) transition temperature, TCDWT_{\mathrm{CDW}}. Noticeably, this magnetic above TCDWT_{\mathrm{CDW}} does not follow the Curie-Weiss paramagnetic characteristic as its sister compound YV6Sn6\mathrm{YV}_{6}\mathrm{Sn}_{6} [3] follows. After 50 K which is far below the TCDWT_{\mathrm{CDW}}, the magnetization starts to increase with decreasing temperature and behaves like a week Pauli paramagnetic behavior [4]. This is also a distinct behavior from the well-known AV3Sb5\mathrm{AV}_{3}\mathrm{Sb}_{5} (A=\mathrm{A}= K, Rb, Cs) family, which follows the Curie-Weiss law with a small effective moment (0.22 μB\mu_{\mathrm{B}} / V atom for KV3Sb5\mathrm{KV}_{3}\mathrm{Sb}_{5}) [5].

From Figure S2, the field dependent magnetization varies linearly at all measured temperatures and does not show any nonlinear behavior. When magnetic field is applied along z^\hat{z}, as shown in Figure S1 and (Figure S1), the magnetization is slightly larger than along x^\hat{x} but follows almost the same behavior. Remarkably, the dip induced by the magnetic field is observed at about 250 K when the magnetic field is aligned along z^\hat{z}. However, the reason for the dip is still unclear and may need further attention.

Additional electronic transport measurements and analyses

Figure S3 shows the additional measurement of electronic transport properties when magnetic field is applied along x^\hat{x} and electric current is injected along y^\hat{y}. The temperature dependent resistivity shows a sharp drop at around TCDWT_{\mathrm{CDW}} and follows almost the same behavior with different fields as shown in Figure S3. The magnetoresistance (MR = (ρ(B)/ρ(0)−1)×100%(\rho(B)/\rho(0)-1) \times100\%), shown in Figure S3, follows a parabolic dependence in the low magnetic field range, but it becomes linear with further field increase. The concentration and mobility of the electron-type carrier extracted from the Hall resistivity in the high field range show an obvious change around TCDWT_{\mathrm{CDW}}, which may be due to the phonon softening. A dramatic sign change of the carrier concentration, which appears only at the CDW phase transition, implies a strong interaction between charge carrier and lattice dynamics. The values of concentration and mobility at 2K extracted from the high field slope of the Hall resistivity are 2.36×1021 cm−32.36 \times10^{21}\ \mathrm{cm}^{-3} and 265 cm2 V−1 s−1265\ \mathrm{cm}^{2}\ \mathrm{V}^{-1}\ \mathrm{s}^{-1}, respectively, as shown in Figure S3. The charge carriers decrease slightly with increasing temperature, while the mobility decreases more rapidly with increasing temperature up to the CDW transition. Figure S3 shows the Hall resistivity as function of the rotating angle from B∥x^B \parallel\hat{x} to B∥y^B \parallel\hat{y}, i.e., from being perpendicular to the current to parallel to the current, which also shows the anomalous Hall effect (AHE) like non-linear behavior.

Then we changed the current direction and now the current is along z^\hat{z} but the field is the same as along x^\hat{x}. The measured data are displayed in Figure S4 and show similar behavior to the I∥y^I \parallel\hat{y} and B∥x^B \parallel\hat{x} measurement geometry, differing only in magnitude. For example, the ρzz\rho_{zz} shows a stronger dependence on the magnetic field, electron type charge carrier as expected from the electronic band structure. Additionally, the ρzz\rho_{zz} (Figure S4) as a function of angle dependent at various magnetic fields have an obvious twofold symmetry when the field is rotated in a full circle, following a conventional trigonometric dependence. From our results, the magnetization is almost negligible in ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} as shown in Figure S2. However, the Hall measurements exhibit a similar feature of AHE, which is seen up to the CDW phase transition temperature.

Resistivity, anomalous Hall, carrier density and mobility, angular resistivity

Figure S4. Resistivity, anomalous Hall, carrier density and mobility, angular resistivity. (a) Temperature-dependent resistivity ρzz\rho_{zz} when I∥z^I \parallel\hat{z}, B∥x^B \parallel\hat{x} at different magnetic fields. (b) Hall resistivity ρyz\rho_{yz} as functions of magnetic field at different temperatures. (c) Estimated carrier concentration and mobility as functions of temperature. (d) Angle-dependent longitudinal resistivity ρzz\rho_{zz} at 2 K for several magnetic fields. The red dashed lines are cosine function fit. The inset shows that the field is rotated from x^\hat{x} to z^\hat{z}.

A nonlinear Hall signal in nonmagnetic systems is primarily attributed to the presence of multiple bands. Such Hall is described by the standard two-band model, as

ρzy(B)=1eμ12μ22(n1+n2)B3+(μ12n1+μ22n2)Bμ12μ22(n1+n2)2B2+(μ1n1+μ2n2)2(1)\rho_{zy}(B)=\frac{1}{e}\frac{\mu_{1}^{2}\mu_{2}^{2}(n_{1}+n_{2})B^{3}+(\mu_{1}^{2}n_{1}+\mu_{2}^{2}n_{2})B}{\mu_{1}^{2}\mu_{2}^{2}(n_{1}+n_{2})^{2}B^{2}+(\mu_{1}n_{1}+\mu_{2}n_{2})^{2}} \tag*{(1)}

where μ1,μ2\mu_{1},\mu_{2} are the mobilities and n1,n2n_{1},n_{2} are the density of the carriers, respectively. For a more reliable analysis, the Hall conductivity, σyz=ρzyρzzρyy+ρzyρyz≈ρzyρzzρyy\sigma_{yz}=\frac{\rho_{zy}}{\rho_{zz}\rho_{yy}+\rho_{zy}\rho_{yz}}\approx\frac{\rho_{zy}}{\rho_{zz}\rho_{yy}} and the longitudinal conductivity, σyy=ρzzρzzρyy+ρzyρyz≈1ρyy\sigma_{yy}=\frac{\rho_{zz}}{\rho_{zz}\rho_{yy}+\rho_{zy}\rho_{yz}}\approx\frac{1}{\rho_{yy}}, when ρzzρyy≫ρzyρyz\rho_{zz}\rho_{yy}\gg\rho_{zy}\rho_{yz}, are estimated from the measured Hall and longitudinal resistivities. Then, the following conventional relations of the two-band model for Hall and longitudinal conductivities are fitted [6, 7, 8]:

σyz(B)=eB(n1μ121+μ12B2+n2μ221+μ22B2)(2)\sigma_{yz}(B)=eB\left(\frac{n_{1}\mu_{1}^{2}}{1+\mu_{1}^{2}B^{2}}+\frac{n_{2}\mu_{2}^{2}}{1+\mu_{2}^{2}B^{2}}\right) \tag*{(2)}
σyy(B)=e(n1μ11+μ12B2+n2μ21+μ22B2).(3)\sigma_{yy}(B)=e\left(\frac{n_{1}\mu_{1}}{1+\mu_{1}^{2}B^{2}}+\frac{n_{2}\mu_{2}}{1+\mu_{2}^{2}B^{2}}\right). \tag*{(3)}

The fitted results of σyz\sigma_{yz} and σyy\sigma_{yy} are summarized in Figure S5. From Figure S5 and (b), the σyz\sigma_{yz} and σyy\sigma_{yy} are well fitted using Eq. (2) and Eq. (3), respectively. However, the fitted parameters are completely different from each other: the fit of σyz\sigma_{yz} indicates electron carriers only, while the fit of σyy\sigma_{yy} gives both electron and hole carriers, indicating a clear disparity between these two fits. Usually, both fits reflect the same parameters, conforming the applicability of the two-band model, as reported in several typical multiband systems [8, 9]. We also examined the σyz\sigma_{yz} and σyy\sigma_{yy} fits by considering the same initial parameters from Figure S5 and (b) in the two-band model. In this way, the fits are the worst and do not converge as shown in Figure 5 (c) and (d). Therefore, the two-band model may not be the correct mechanism to describe the nonlinear Hall behavior in ScV6_6Sn6_6.

Additional Shubnikov de-Haas oscillations analyses

The other important physical property, the Shubnikov–de Haas (SdH) quantum oscillation, is also observed in the high-quality ScV6Sn6\mathrm{ScV}_{6}\mathrm{Sn}_{6} crystals, as shown in Figure S6. We present the first derivative of ρzz\rho_{zz} in Figure S6 to show the SdH oscillation more clearly. The large magnetoresistance (170% at 2 K and 9 T) and the SdH oscillation indicate the high quality of our crystals. Figure S6 shows the frequency dependence calculated by Fast Fourier Transformation (FFT), which gives three frequencies at 10, 28, and 50 T labeled as α\alpha, β\beta, and γ\gamma with respect to three Fermi pockets as discussed in the main text. Due to the multiple frequency as we mentioned in the main text, it is difficult to estimate the parameters related to the Fermi surfaces (FSs) from the whole FFT spectrum. Thus, we extracted the frequency components for further analysis by using bandwidth pass as shown in Figure S6 [10]. However, by using FFT filtering process the oscillation decay as increasing magnetic field (i.e. decreasing 1/B1/B), as shown in Figure S6 and Figure S7 for reconstructed β\beta- and γ\gamma-pocket, respectively, suggesting the presence of beating patterns which influence the Berry’s phase analysis [11, 12]. Thus, the extracted result is relatively unreliable. We also directly analyzed the angle dependent SdH curves for some curves that have only one frequency, as shown in Figure S7 – Figure S7. the estimated Berry’s phases are listed and discussed in Figure 2g in the main text. The so-called Dingle plot is also presented in Figure S8 for γ\gamma pocket when magnetic field is parallel to z^\hat{z}-axis, giving TD=2.43T_D = 2.43 K and τq=5×1013\tau_q = 5 \times10^{13} s for FγF_\gamma. The other estimated parameters of both α\alpha, β\beta, and γ\gamma FS pockets are listed in Table S1.

SdH oscillations and Landau level fan diagrams

Figure S7. SdH oscillations and Landau level (LL) fan diagram of γ\gamma-pocket at 2 K for various rotating angles (BB rotates from x^\hat{x} to z^\hat{z}). (a-d) are the data of θ=60∘\theta= 60^\circ, 70∘70^\circ, 80∘80^\circ, 90∘90^\circ, respectively. The LLs are assigned directly from the observed oscillations without filtering.

The magnetoresistance (MR) for the rotating magnetic field from x^\hat{x} to z^\hat{z} and x^\hat{x} to y^\hat{y} are shown in Figure S9. At T=2T = 2 K, the oscillation of Δρzz\Delta\rho_{zz} shows an obvious change when the field is rotated in the x^\hat{x} - z^\hat{z} plane, illustrating an anisotropic FS. Contrastingly, when magnetic field is rotating in x^\hat{x} - y^\hat{y} plane, the oscillation spectrum is almost the same, indicating that the FS is almost isotropic in this plane.

Electronic band structure calculations

Brillouin zones of ScV3Sn6

Figure S10. Brillouin zones (BZs) of ScV3Sn6ScV_{3}Sn_{6}. (a) Three-dimensional (3D) BZ of high-temperature (280 K) phase of ScV6Sn6ScV_{6}Sn_{6} with space group P6/mmmP6/mmm. (b) 3D BZ of low-temperature (50 K) phase of ScV6Sn6ScV_{6}Sn_{6} with space group R32R32. Crystallographic information used in calculation are taken from the previous report [1].

We have calculated the band structures of ScV6Sn6ScV_{6}Sn_{6} in the high temperature (HT) phase as well as in the low temperature (LT) CDW phase. The LT phase band structure is unfolded to match the one of HT phase when considering spin orbital coupling (SOC) (Figure S11) and neglecting SOC (Figure S11), respectively. The band structures didn’t change a lot with or without considering SOC. Specifically, for LT phase, energy bands along AA-HH-LL-AA (i.e. kz=πk_z = \pi plane) are gapped by CDW phase transition, which is consist with the results of angle-resolved photoemission spectrum (ARPES) [13]. The orbital projected band structure is also calculated and plotted in Figure S11 and d. It is clear that the kagome V dd orbitals dominate the bands close to the Fermi level. Only a few bands along MM-KK, GG-AA, and HH-KK are contributed by Sn pp orbitals, while the dd orbitals from Sc are negligible.

References

  1. [1]H. W. S. Aracchige, W. R. Meier, M. Marshall, T. Matsuoka, R. Xue, M. A. McGuire, R. P. Hermann, H. Cao, and D. Mandrus, Charge Density Wave in Kagome Lattice Intermetallic ScV₆Sn₆, Phys. Rev. Lett. 129, 216402 (2022).
  2. [2]J. Rodríguez-Carvajal, Recent Advances in Magnetic Structure Determination by Neutron Powder Diffraction, Physica B: Condensed Matter 192, 55 (1993).
  3. [3]G. Pokharel, S. M. L. Teicher, B. R. Ortiz, P. M. Sarte, G. Wu, S. Peng, J. He, R. Seshadri, and S. D. Wilson, Electronic Properties of the Topological Kagome Metals YV₆Sn₆ and GdV₆Sn₆, Phys. Rev. B 104, 235139 (2021).
  4. [4]S. Blundell, Magnetism in Condensed Matter, Reprint (Oxford Univ. Press, Oxford, 2014).
  5. [5]B. R. Ortiz, L. C. Gomes, J. R. Morey, M. Winiarski, M. Bordelon, J. S. Mangum, I. W. H. Oswald, J. A. Rodriguez-Rivera, J. R. Neilson, S. D. Wilson, et al., New Kagome Prototype Materials: Discovery of KV₃Sb₅, RbV₃Sb₅, and CsV₃Sb₅, Phys. Rev. Materials 3, 094407 (2019).
  6. [6]N. Bansal, Y. S. Kim, M. Brahlek, E. Edrey, and S. Oh, Thickness-Independent Transport Channels in Topological Insulator Bi₂Se₃ Thin Films, Phys. Rev. Lett. 109, 116804 (2012).pubmed.ncbi.nlm.nih.gov/23005664
  7. [7]F. C. Chen, Y. Fei, S. J. Li, Q. Wang, X. Luo, J. Yan, W. J. Lu, P. Tong, W. H. Song, X. B. Zhu, et al., Temperature-Induced Lifshitz Transition and Possible Excitonic Instability in ZrSiSe, Phys. Rev. Lett. 124, 236601 (2020).DOI
  8. [8]J. Xu, Y. Wang, S. E. Pate, Y. Zhu, Z. Mao, X. Zhang, X. Zhou, U. Welp, W.-K. Kwok, D. Y. Chung, et al., Unreliability of Two-Band Model Analysis of Magnetoresistivities in Unveiling Temperature-Driven Lifshitz Transition, Phys. Rev. B 107, 035104 (2023).DOI
  9. [9]Y. L. Wang, L. R. Thoutam, Z. L. Xiao, J. Hu, S. Das, Z. Q. Mao, J. Wei, R. Divan, A. Luican-Mayer, G. W. Crabtree, et al., Origin of the Turn-on Temperature Behavior in WTe₂, Phys. Rev. B 92, 180402 (2015).DOI
  10. [10]M. N. Ali, L. M. Schoop, C. Garg, J. M. Lippmann, E. Lara, B. Lotsch, and S. S. P. Parkin, Butterfly Magnetoresistance, Quasi-2D Dirac Fermi Surface and Topological Phase Transition in ZrSiS, Sci. Adv. 2, e1601742 (2016).DOI
  11. [11]X. Xu, X. Wang, T. A. Cochran, D. S. Sanchez, G. Chang, I. Belopol ski, G. Wang, Y. Liu, H.-J. Tien, X. Gui, et al., Crystal Growth and Quantum Oscillations in the Topological Chiral Semimetal CoSi, Phys. Rev. B 100, 045104 (2019).DOI
  12. [12]N. Huber, K. Alpin, G. L. Causer, L. Worch, A. Bauer, G. Benka, M. M. Hirschmann, A. P. Schnyder, C. Pfleiderer, and M. A. Wilde, Network of Topological Nodal Planes, Multifold Degeneracies, and Weyl Points in CoSi, Phys. Rev. Lett. 129, 026401 (2022).DOI
  13. [13]S. Lee, C. Won, J. Kim, J. Yoo, S. Park, J. Denlinger, C. Jozwiak, A. Bostwick, E. Rotenberg, R. Comin, et al., Nature of Charge Density Wave in Kagome Metal ScV₆Sn₆, arXiv:2304.11820v1 (2023).arxiv.org/abs/2304.11820

Paper details

vlnc:W_unnp3nc4
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 9

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