---
title: 'ScV6Sn6: Kagome Metal & CDW Physics'
url: https://www.emergentmind.com/topics/scv6sn6
type: topic
---

# ScV6Sn6: Kagome Metal & CDW Physics

Searching arXiv for recent papers on ScV6Sn6 to ground the article in the latest literature.
ScV\(_6\)Sn\(_6\) is a vanadium-based bilayer kagome metal in the “166” family that crystallizes in the hexagonal \(P6/mmm\) structure and undergoes a first-order charge-density-wave (CDW) transition near \(92\)–\(97\) K. Across the literature, it is characterized by a \(\sqrt{3}\times\sqrt{3}\times 3\) structural modulation with ordering vector \(\mathbf{q}=(1/3,1/3,1/3)\), pronounced reconstruction of the Fermi surface, and kagome-derived electronic states including Dirac crossings, van Hove singularities, and flat-band features. A central theme of current research is that, unlike several other kagome metals, the CDW in ScV\(_6\)Sn\(_6\) is predominantly tied to structural and phononic instabilities involving Sc and Sn sublattices, although the reconstructed low-temperature state still hosts quantum-oscillation signatures of non-trivial band topology [2205.04582].

## 1. Crystal framework and kagome-derived electronic structure

At room temperature ScV\(_6\)Sn\(_6\) crystallizes in hexagonal \(P6/mmm\). Multiple studies describe it as a bilayer kagome system in which V atoms form two parallel kagome layers per unit cell, interleaved with Sc and Sn spacer layers. Reported lattice parameters are closely consistent across measurements and calculations, with representative values including \(a=5.4730\) Å and \(c=9.1719\) Å at room temperature, \(a=b=5.4733\) Å and \(c=9.1724\) Å, and \(a=b=5.456\) Å and \(c=9.230\) Å in related datasets. The structural motif is variously described as stacked Sc–Sn(1) layers, hexagonal Sn(2) layers, and two V–Sn kagome nets, or as a repeating sequence in which V kagome nets are separated by ScSn\(_2\) and Sn\(_2\) layers [2403.02463].

The electronic structure exhibits the canonical kagome features. Density-functional and photoemission studies place Dirac crossings at \(K\), van Hove singularities at \(M\), and nearly flat bands near the Fermi level. One \(\mu\)-ARPES study locates the Dirac cone at \(K\) at \(E\approx -0.21\) eV and identifies two van Hove singularities at \(M\), one from out-of-plane V \(d_{z^2}\) and another from in-plane V \(d_{x^2-y^2}/d_{xy}\) orbitals. Related ARPES work reports three van Hove singularities near \(E_F\), two within \(\pm 30\) meV of \(E_F\), while another STM/QPI study resolves a peak in \(dI/dV\) at \(E_p\approx +14\) meV associated with a saddle point. The low-energy states are predominantly V \(3d\)-derived, although a three-dimensional \(\Gamma\)-centered band with dominant planar Sn character is also important for the ordered state [2402.03765].

Below the ordering temperature, the lattice reconstructs into a \(\sqrt{3}\times\sqrt{3}\times 3\) supercell. Real-space descriptions emphasize predominantly out-of-plane displacements of Sc and Sn atoms and only weak buckling or very small in-plane shifts of the V kagome network. Representative refined amplitudes are \(A_{\mathrm{Sc}}\simeq 0.16\) Å and \(A_{\mathrm{Sn1}}\simeq 0.16\) Å along \(c\), whereas V shifts are of order \(0.004\)–\(0.006\) Å. This hierarchy of displacements is central to why the CDW is widely interpreted as structurally driven rather than a direct instability of the V kagome sublattice [2205.04582].

## 2. Charge-density-wave transition and competing ordering vectors

The primary low-temperature transition occurs near \(T_{\mathrm{CDW}}\approx 92\)–\(97\) K and is consistently described as first-order or weakly first-order. Thermodynamic and transport signatures include a sharp resistive anomaly, thermal hysteresis of order \(0.3\)–\(0.5\) K, a narrow heat-capacity peak, abrupt changes in optical reflectivity, and a sudden onset of superlattice diffraction intensity. In one early diffraction refinement, the order parameter was expressed through the CDW diffraction intensity as \(I(T)\propto |\psi(T)|^2\), with \(\psi(T)\) jumping to a finite value at \(T_{\mathrm{CDW}}\) [2211.03412].

The established long-range ordering vector is
\[
\mathbf{q}=\left(\tfrac13,\tfrac13,\tfrac13\right),
\]
corresponding to a \(\sqrt{3}\times\sqrt{3}\) in-plane supercell and tripling along \(c\). However, a substantial literature now shows that this ordered state emerges from a more complex precursor landscape. Synchrotron diffraction identifies short-range correlations at \(\mathbf{q}_{1/2}=(1/3,1/3,1/2)\) above the transition, with an inter-layer correlation length that grows on cooling and then is interrupted when long-range order freezes in at \(\mathbf{q}_{1/3}=(1/3,1/3,1/3)\). In that study, the out-of-plane correlation length of the frustrated mode follows
\[
\xi_\perp(T)\propto |T-T_{\mathrm{CO}}|^{-\nu},
\]
and reaches approximately \(3c\) near \(T_{\mathrm{CO}}\), at which point the system undergoes a first-order jump into the long-range \((1/3,1/3,1/3)\) state. Disorder leaves remnant \((1/3,1/3,1/2)\) diffuse scattering to the lowest measured temperatures, suggesting incomplete conversion of all chain segments [2307.11843].

A related first-principles perspective describes the CDW as arising from a Jahn–Teller-like interlayer Sn–Sn dimerization. In that account, a \(\sqrt{3}\times\sqrt{3}\times 2\) pattern associated with \(H=(1/3,1/3,1/2)\) is energetically favorable at low temperature, whereas the experimentally observed \(\sqrt{3}\times\sqrt{3}\times 3\) stacking becomes thermodynamically stabilized at finite temperature by configurational entropy. The reported configurational contribution is
\[
\Delta S_{\mathrm{conf}}\simeq k_B \ln 6 / 3 = 0.051\ \mathrm{meV\cdot K^{-1}\cdot f.u.^{-1}},
\]
and the calculated free-energy crossing occurs near \(97\) K, close to experiment. This suggests an order–disorder component and dynamic fluctuations among nearly degenerate stackings above the static transition [2308.13796].

These two strands of evidence converge on a picture in which ScV\(_6\)Sn\(_6\) is not a simple single-\(\mathbf{q}\) CDW system. Rather, the observed long-range order is selected from a frustrated manifold involving closely related stacking vectors and cooperative distortions of the Sn–Sc–Sn structural motifs.

## 3. Microscopic origin of the ordered state

A major result of the ScV\(_6\)Sn\(_6\) literature is that the CDW is not well described by a conventional Fermi-surface-nesting instability of the kagome-derived van Hove singularities. Optical spectroscopy reports no observable density-wave-like gap opening behavior across the transition: neither \(R(\omega)\) nor \(\sigma_1(\omega)\) shows the gradual low-energy depletion and coherence peak expected for a standard nesting-driven CDW. Instead, the data show an abrupt spectral reconstruction, consistent with a first-order structural transition and sudden band-structure reorganization [2211.03412].

ARPES further sharpens this distinction. One study finds that the V-derived van Hove singularities remain unshifted and ungapped across \(T_{\mathrm{CO}}\), while a three-dimensional \(\Gamma\)-centered band with dominant planar Sn character opens a large gap
\[
\Delta_{\mathrm{CDW}} \approx 260\ \mathrm{meV}
\]
and reconstructs the Fermi surface into a star-shaped pattern. That work emphasizes that the transition remains robust in Sc\((\mathrm{V}_{1-x}\mathrm{Cr}_x)_6\)Sn\(_6\) for \(0\le x\le 0.10\), corresponding under a rigid-band approximation to a Fermi-level shift \(\Delta E_F\approx 120\) meV, which argues against a van-Hove-tuned electronic mechanism [2304.11820].

A closely related \(\mu\)-ARPES and first-principles comparison between ScV\(_6\)Sn\(_6\) and non-CDW YV\(_6\)Sn\(_6\) reaches the same conclusion by a different route. The joint density of states,
\[
C(q)=\frac{1}{(2\pi)^3}\int A(k,E_F)\,A(k+q,E_F)\,d^2k,
\]
shows only a trivial peak at \(q=0\), with no enhancement at \(Q_{\mathrm{CDW}}=(1/3,1/3,1/3)\) in either compound. At the same time, CDW-folded bands in ScV\(_6\)Sn\(_6\) exhibit multiple hybridization gaps, including \(\Delta_2\simeq 70\) meV and \(\Delta_3\simeq 40\) meV, together with a reproducible kink near \(35\) meV that was interpreted as a hallmark of electron–boson coupling. First-principles phonons identify soft low-energy modes of the quasi-one-dimensional Sc–Sn chains along \(c\), including an imaginary branch absent in YV\(_6\)Sn\(_6\) [2402.03765].

Time-domain experiments add dynamical evidence. Ultrafast optical spectroscopy resolves a coherent amplitude mode near \(1.45\) THz and shows that the electronic subsystem saturates at fluence \(F_{\mathrm{sat}}\approx 240\ \mu\mathrm{J/cm^2}\), while the lattice CDW survives up to \(\sim 700\ \mu\mathrm{J/cm^2}\) and melts only near \(1000\ \mu\mathrm{J/cm^2}\). A later time-resolved X-ray study reports sub-200 fs CDW suppression with \(t_m\simeq 160\) fs, a coherent amplitudon at \(f=1.48\) THz, and a DFT-derived double-well free-energy landscape with an energy barrier of order \(\sim 10\) meV. Both studies interpret these observations as hallmarks of a phonon-coupled charge order [2506.14888].

Taken together, these results establish that the structural degrees of freedom of Sc and Sn are primary, while kagome-derived electronic states are strongly reconstructed but are not the dominant driver of the transition.

## 4. Fermi-surface reconstruction, optical response, and local probes

The CDW reconstructs the Fermi surface severely. In one DFT treatment, the pristine phase has two bands crossing \(E_F\): a strongly three-dimensional sheet with necks along \(\Gamma\)–\(A\) and small nearly two-dimensional electron pockets at the Brillouin-zone boundary. In the CDW state, the Fermi surface reorganizes into four sheets, including prismatic and cylindrical hole pockets centered on \(\Gamma\), an outer quasi-two-dimensional cylinder, and tiny electron pockets near \(K\). This reconstruction is the basis for interpreting the quantum-oscillation spectrum in the ordered phase [2310.00751].

Optically, the first-order band reorganization is visible as an abrupt transfer of Drude weight into interband structures. Below the transition, part of the Drude spectral weight reappears as Lorentz-type peaks at approximately \(570\ \mathrm{cm^{-1}}\) (\(0.071\) eV), \(2460\ \mathrm{cm^{-1}}\) (\(0.305\) eV), and \(3780\ \mathrm{cm^{-1}}\) (\(0.469\) eV). Simultaneously, the plasma frequency drops from \(\simeq 15200\ \mathrm{cm^{-1}}\) at \(300\) K to \(\simeq 11900\ \mathrm{cm^{-1}}\) at \(5\) K, while the scattering rate collapses from \(\simeq 630\ \mathrm{cm^{-1}}\) to \(\simeq 60\ \mathrm{cm^{-1}}\). This is consistent with a metallic low-temperature state with fewer itinerant carriers and much lower dissipation [2211.03412].

Local probes show that the ordered state remains only partially gapped. STM/STS on the kagome termination reveals a partial gap of about \(20\) meV at the Fermi level and a \(\sqrt{3}\times\sqrt{3}R30^\circ\) modulation, whereas the Sn\(_1\) termination shows neither the modulation nor the gap. The ratio
\[
\frac{2\Delta}{k_B T^*}\approx 5
\]
is substantially smaller than the \(\sim 10\)–\(11\) reported for AV\(_3\)Sb\(_5\) and FeGe, which was interpreted there as evidence for weaker coupling. Temperature-dependent ARPES in the same work finds abrupt spectral-weight redistribution below \(T^*\), consistent with the first-order character of the transition [2302.12227].

\(^{51}\)V NMR provides a complementary bulk local picture. Below the transition, the quadrupole pattern splits into three distinct environments with \(\nu_Q\approx 463\), \(489\), and \(499\) kHz, directly reflecting the commensurate modulation with \(q=(1/3,1/3,1/3)\). The spin-lattice relaxation rate drops by approximately a factor of two, and DFT yields
\[
\frac{D_{\mathrm{CDW}}(E_F)}{D_{\mathrm{HT}}(E_F)}=\frac{1}{\sqrt{2}},
\]
which quantitatively accounts for the observed change through the Korringa relation. The same NMR study reports orientation-dependent splitting patterns consistent with orbital-selective modulations of the local density of states, with different behavior for \(d_{x^2-y^2}/d_{xy}\) and \(d_{z^2}\) orbitals [2404.18597].

A plausible implication is that the low-temperature electronic structure is best understood not as uniformly gapped, but as an orbitally selective and strongly momentum-dependent reconstruction of a still-metallic state.

## 5. Quantum oscillations and topological bands

Quantum-oscillation studies establish that ScV\(_6\)Sn\(_6\) hosts light carriers and at least one orbit with a non-trivial Berry phase. High-field torque magnetometry up to \(43\) T reveals six principal de Haas–van Alphen frequencies in the CDW phase:
\[
F_\alpha=77\pm 12\ \mathrm{T},\ 
F_\gamma=155\pm 8\ \mathrm{T},\ 
F_\chi=235\pm 10\ \mathrm{T},\ 
F_\delta=314\pm 15\ \mathrm{T},\ 
F_\varepsilon=394\pm 18\ \mathrm{T},\ 
F_\beta=2834\pm 30\ \mathrm{T},
\]
with a faint shoulder at \(\simeq 2960\) T. Fits to the Lifshitz–Kosevich expression give \(m^*_\alpha=0.19\,m_e\) and \(m^*_\beta=0.77\,m_e\), while Dingle temperatures \(T_D\lesssim 10\) K indicate high crystal quality. A Landau-level fan analysis of the \(\alpha\) oscillation yields intercepts \(\gamma=\Phi_B/2\pi\simeq 0.37\)–\(0.43\), corresponding to \(\Phi_B\approx \pi\) [2310.00751].

Transport-based Shubnikov–de Haas measurements in a lower-field range detect smaller frequencies and also report a non-trivial phase. One dataset gives \(F_\alpha=10\) T, \(F_\beta=28\) T, and \(F_\gamma=50\) T, with effective masses \(m^*_\alpha=0.105\,m_0\), \(m^*_\beta=0.128\,m_0\), and \(m^*_\gamma=0.180\,m_0\). From a Landau fan for the \(50\) T orbit, the intercept is \((\gamma-\delta)=-0.101\pm 0.005\), yielding \(\phi_B\simeq \pi\) for a two-dimensional assignment. The same study interprets this \(\gamma\)-pocket as a Dirac-derived Fermi-surface sheet that survives the CDW with only a small SOC gap of order \(10\) meV [2305.04683].

A later combined SdH/dHvA work refines the pocket assignments in the low-temperature folded phase. It reports \(F_\alpha\simeq 21\) T and \(F_\beta\simeq 44\) T in transport, with \(m^*_\alpha=0.12\,m_e\) and \(m^*_\beta=0.19\,m_e\), while dHvA yields \(m^*_\alpha\simeq 0.15\,m_e\) and \(m^*_\beta\simeq 0.23\,m_e\). For the \(\beta\) orbit, \(n_0(\beta)\simeq 0.43\) implies \(\phi_B(\beta)\simeq 1.11\pi\simeq \pi\), and multicomponent LK fits as well as torque histograms support a robust non-trivial phase. That work identifies the \(\beta\) orbit with a small ellipsoidal pocket centered at \(K_1=(1/3,1/3,1/3)\) and argues that it encloses a Dirac node [2409.05634].

Across these studies, the details of the resolved frequencies vary with field window, measurement channel, and sample condition, but the recurring result is consistent: the CDW-reconstructed state retains at least one light, Dirac-derived orbit with Berry phase approximately \(\pi\).

## 6. Symmetry, transport anomalies, and the evolving phase diagram

The symmetry properties of the CDW state remain an active subject because different probes emphasize different aspects of the ordered phase. Several transport studies reported Hall anomalies resembling an anomalous Hall effect. Under pressure, the Hall resistivity in the parent compound shows a steep low-field rise and saturation below \(T_{\mathrm{CDW}}\), while Cr-substituted samples are instead well described by a standard two-band model. In the parent under pressure, the anomalous Hall-like component disappears sharply at \(T_{\mathrm{CDW}}(P)\), and the scaling \(\sigma_{xy}^A\propto \sigma_{xx}^2\) was interpreted as an “intrinsic”-like regime with a superquadratic slope [2403.02463].

However, a much more recent study on very clean crystals argues that the low-field Hall anomaly is quantitatively inconsistent with an intrinsic anomalous Hall effect. By increasing the in-plane residual-resistivity ratio from \(\lesssim 3\) to \(>50\), that work shows that the weak-field nonlinearity in \(\rho_{xy}(B)\) becomes steeper and moves to lower fields as mobility increases, exactly as expected from ordinary multiband transport. A conductivity model
\[
\sigma_{xx}(B)= e\sum_i \frac{n_i\mu_i}{1+(\mu_i B)^2},\qquad
\sigma_{xy}(B)= eB\sum_i \frac{n_i\mu_i^2}{1+(\mu_i B)^2}
\]
captures the behavior, while a previously reported “plateau” disappears in cleaner crystals. In parallel, the same work resolves up to six closely spaced anomalies near \(90\) K in the cleanest samples and uses elastoresistivity to show that an intermediate phase breaks three-fold rotational symmetry [2606.29048].

Independent evidence for an intermediate nematic state has also been reported by transport, optical polarization rotation, calorimetry, and Sagnac interferometry. In that account, \(T_{\mathrm{CDW}}\approx 93.0\) K marks a \(3\times 3\times 1\) CDW, and a second transition at \(T^*\approx 86.0\) K produces a spontaneous in-plane \(C_6\to C_2\) anisotropy without time-reversal-symmetry breaking. The optical polarization rotation follows
\[
\phi_T(\alpha)=\phi_p\sin(2\alpha-\alpha_0),
\]
with amplitude up to \(\pm 0.5\) mrad, while the zero-field Kerr signal remains within \(\pm 0.02\ \mu\mathrm{rad}\) [2502.14568].

A lower-temperature electronic nematicity has been identified by STM/STS, which observes stripe-like intra-unit-cell order, anisotropic annihilation of two of the three van Hove singularities, and an elliptical QPI ring. In that study, the inferred eccentricity is \(e\approx 0.72\), and the phenomenology is interpreted in terms of a Pomeranchuk instability on the kagome lattice [2406.13702].

Time-reversal symmetry breaking is correspondingly controversial. One Kerr study finds no spontaneous polar Kerr effect to within \(\sim 100\) nrad, even after field training and under \(\pm 0.1\%\) uniaxial strain, concluding that the CDW phase preserves time-reversal symmetry within that sensitivity [2510.26010]. NMR likewise reports no extra broadening or line shifts indicative of static TRS breaking on the NMR timescale [2404.18597]. By contrast, a later anomalous Nernst and \(\mu\)SR study reports an anomalous transverse Nernst signal reaching \(\sim 4\ \mu\mathrm{V/K}\) near \(T_{\mathrm{CDW}}\) and interprets the combined transport and \(\mu\)SR data as evidence for TRS breaking, while explicitly noting that hidden magnetism, orbital currents, and chiral charge order remain competing possibilities [2508.12667].

The current literature therefore supports a cautious synthesis. ScV\(_6\)Sn\(_6\) clearly hosts a first-order CDW with a rich nearby symmetry landscape, including nematic responses and multiple closely spaced transitions in clean samples. Whether the low-field transverse anomalies require intrinsic TRS breaking, or can be fully reduced to multiband transport plus symmetry-lowered electronic structure, remains unsettled.

## 7. Tunability and broader significance within kagome materials

ScV\(_6\)Sn\(_6\) is highly tunable by disorder, chemical substitution, pressure, and strain. Substituting \(5\%\) Cr at the V site or applying \(1.5\) GPa pressure shifts the CDW from \(92\) K to about \(50\)–\(54\) K, and pressure of \(P\gtrsim 2.5\) GPa fully suppresses the CDW in one transport study. Cr substitution changes the magnetoresistance from quasi-linear to quadratic, whereas under pressure the quasi-linear response persists until the CDW is suppressed. These contrasting effects were attributed to how substitution and pressure differently modify phonons, disorder, and carrier compensation [2403.02463].

Rare-earth substitution on the Sc site also strongly affects the order. Light Y substitution rapidly suppresses the \((1/3,1/3,1/3)\) long-range order and then the \((1/3,1/3,1/2)\) precursor correlations, consistent with a scenario in which larger ions locally stiffen the trimer centers and quench the movable-chain instability [2307.11843]. By contrast, comparison with YV\(_6\)Sn\(_6\) as a stoichiometric endpoint shows similar fermiology but no CDW, underscoring the sensitivity of the instability to subtle structural and phononic details rather than gross band topology alone [2402.03765].

Uniaxial strain has emerged as a qualitatively different tuning parameter. High-resolution X-ray diffraction under in-plane compression finds that anisotropic strain stabilizes and enhances the CDW rather than suppressing it. Compression along \([H00]\) and \([HH0]\) lowers the symmetry from hexagonal to orthorhombic, selects a single in-plane CDW domain, and increases \(T_{\mathrm{CDW}}\) with slopes
\[
\partial T_{\mathrm{CDW}}/\partial \varepsilon_{[H00]}\simeq -23\ \mathrm{K/\%},\qquad
\partial T_{\mathrm{CDW}}/\partial \varepsilon_{[HH0]}\simeq -18\ \mathrm{K/\%}.
\]
That work argues that the enhancement comes primarily from strain-selective ordering of frustrated Sn\(^T\)–Sc–Sn\(^T\) rattling chains, not from large hardening of the imaginary phonon branch, and models the effect within a strain-dependent three-state Potts framework [2606.05866].

This body of work positions ScV\(_6\)Sn\(_6\) as a distinct member of the kagome-metal landscape. It combines kagome-derived Dirac and van Hove physics with a predominantly phonon-driven, three-dimensionally modulated CDW, severe but incomplete Fermi-surface reconstruction, and an unusual proximity to nematic and possibly other hidden symmetry-broken states. Relative to AV\(_3\)Sb\(_5\), where CDW formation is more closely tied to V-sublattice electronic structure and superconductivity emerges at low temperature, ScV\(_6\)Sn\(_6\) offers a complementary platform in which non-kagome structural degrees of freedom, frustrated stacking competition, and topological pockets coexist in a clean metallic environment [2302.10699].

Source: https://www.emergentmind.com/topics/scv6sn6