betaFeedback

Reading v1View latest version

WorksPhysics and Astronomy

Light-Cone Friedel Oscillations: a Window into van Hove Superconductivity

Abstract

Van Hove (vH) singularities can enhance superconductivity by generating a large pairing susceptibility. Yet, it remains challenging to experimentally establish whether pairing really occurs near these points. We show that the saddle-point dispersion near a vH point gives rise to a `light-cone' vH-selective behavior of Friedel oscillations, which is drastically different from the familiar concentric circles. Each vH point maps onto its own real-space light-cone sector, allowing STM around a single impurity to measure whether—and how strongly—that saddle-point quasiparticle sector is gapped. A superconducting gap near vH points suppresses the oscillations in the directions corresponding to peaks in the gap function angular dependence, a behavior that provides a signature of vH-related pairing at STM-accessible length scales. This STM protocol, targeting vH superconductivity, is capable of determining whether superconducting pairing emerges near the vH points or elsewhere on the Fermi surface.

Van Hove (vH) points—saddle points of the electronic dispersion—can strongly influence correlated metals: when the Fermi level lies nearby, their enhanced density of states can amplify interaction-driven instabilities [1, 2, 3, 4, 5, 6, 7, 8, 9]. This enhancement can sharpen the competition among unconventional superconductivity, magnetism, charge or nematic order, and non-Fermi-liquid behavior [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 39, 40]. Here we predict vH-selective Friedel oscillations and show that they provide a unique window into carrier dynamics and pairing near these singular regions of momentum space.

Theory predicts that the enhanced density of states and scattering phase space near vH points strengthen a superconducting pairing channel by concentrating the relevant Cooper-pair susceptibility near the saddle points [9, 18, 19, 20, 21, 22, 23, 24]. This enhancement can produce a momentum-space hot-spot structure in which the superconducting gap is largest at, or close to, the vH points. Yet, this is not a universal correspondence, as the gap structure may also depend on the pairing interaction and the symmetries of the Wannier functions. In special cases, the gap may instead be suppressed at a vH point. Whether superconducting pairing is in fact concentrated in the vH regions is therefore material dependent and must be determined experimentally. Resolving this question requires a probe that can distinguish superconductivity at vH points. Our goal is to provide a practical real-space probe complementary to ARPES, that directly links these two phenomena, by determining whether the quasiparticles associated with individual vH points become gapped by the superconducting order.

With this motivation, here we propose a scanning-tunneling-microscopy (STM) protocol capable of determining whether superconducting pairing is concentrated near the van Hove saddle points, or away from them. Near these points, impurities generate a characteristic vH-selective ‘light-cone’ Friedel oscillations that depend on Minkowski distance, with constant-phase contours

s=(x−xi)2−(y−yi)2=nλF2,λF=2πkF,(1)s = \sqrt{(x-x_i)^2-(y-y_i)^2} = n\frac{\lambda_{\mathrm{F}}}{2}, \qquad\lambda_{\mathrm{F}} = \frac{2\pi}{k_{\mathrm{F}}}, \tag*{(1)}

where (xi,yi)(x_i,y_i) are the impurity coordinates (see also

Refs. [36], [37] and [38].

Oscillatory functions of ss, replacing the conventional circular Friedel oscillations, occur in separate ‘light-cone’ sectors, forming distinct pairs, one per each vH point. Friedel oscillations are confined to vH-point-specific pairs of sectors and decay exponentially in the complementary sectors. In Figure 1 this behavior is illustrated for the twodimensional square lattice bandstructure

Light-cone Friedel oscillations around a point-like impurity in a square-lattice band

Figure 1. (a–d) Light-cone Friedel oscillations around a point-like impurity in a square-lattice band. At small nonzero doping, the interference fringes form hyperbolic patterns Eq. (1), with spacing decreases as 1/∣μ∣1/\sqrt{|\mu|} with Fermi level detuning from the vH point increasing [panels (b) and (c)]. The oscillations are vH-selective: each vH point in momentum space generates hyperbolic fringes confined to a distinct, non-overlapping pair of real-space sectors (XX or YY); in each sector, the hyperbolas asymptote to the sector boundaries. At half filling, μ=0\mu=0, the oscillatory fringes vanish [panel (a)], whereas at larger detuning they evolve into concentric circular fringes [panel (d)]. Shown is the tunneling density of states Eq. (16), calculated for lattice constant a=0.4a=0.4 nm and μ=0\mu=0, −0.2t-0.2t, −0.4t-0.4t, and −3t-3t in panels (a)–(d), respectively. Friedel oscillations arising from vH points are highly sensitive to superconductivity at these points (see Figure 3).

Ek=−2tcos⁡(kxa)−2tcos⁡(kya),(2)E_k = -2t\cos(k_xa)-2t\cos(k_ya), \tag*{(2)}

where vH points occur at the XX and YY symmetry points (π,0)(\pi,0) and (0,π)(0,\pi). At half-filling, localized impurities produce beam-like features in the tunneling current [30, 31, 32, 33, 34, 35]. Slightly away from half-filling, analysis predicts spatial oscillations behaving as

dIdV(r)∼u0cos⁡(2kFs)se−2s/ξ,kF=∣μ∣ta2(3)\frac{dI}{dV}(\mathbf{r}) \sim u_0\frac{\cos(2k_Fs)}{s}e^{-2s/\xi},\qquad k_F=\frac{\sqrt{|\mu|}}{ta^2} \tag*{(3)}

where ξ(Δ,T)\xi(\Delta,T) is the superconducting coherence length given in Eq. (4), μ\mu is Fermi level detuning from vH point and ss is the Minkowski distance given in Eq. (1), taken in the semiclassical limit kFs≫1k_Fs \gg1. The constant phase contours (nodes of cos⁡2kFs\cos2k_Fs) are hyperbolas with periodicity varying with the detuning μ\mu from vH point, as 1/∣μ∣1/\sqrt{|\mu|}. At larger μ\mu the fringes transform into conventional concentric circles. We derive these interference patterns using a continuum saddle-point model with Bogoliubov-de Gennes Green functions and a T-matrix treatment of impurity scattering, and verify them numerically in a realistic tight-binding model on a square lattice.

Our analysis identifies observable signatures of vH points and superconductivity concentrated at these points accessible by state-of-the-art STM measurements. For doping close to vH point, ∣μ∣≪t|\mu| \ll t, the periodicity of the light-cone Friedel oscillations is much greater than atomic spacing, varying with μ\mu as 1/∣μ∣1/\sqrt{|\mu|}. The larger-than-atomic periodicity and a characteristic dependence on detuning from vH point provide clear signatures of the presence of vH points in the bandstructure. A measured suppression of saddle-point fringes in the superconducting state would provide evidence for a gap affecting saddle-point quasiparticles. The decay of light cone oscillations with distance, which is governed by the thermal coherence length ξ(T)=ℏvF/πkBT\xi(T)=\hbar v_F/\pi k_BT in the normal state changes to a shorter lengthscale

ξ(Δ,T)=ℏvFΔ2+π2kB2T2(4)\xi(\Delta,T)=\frac{\hbar v_F}{\sqrt{\Delta^2+\pi^2k_B^2T^2}} \tag*{(4)}

in the superconducting state, a behavior that can be used to determine the value of Δ\Delta at vH point. Since vFv_F scales with vH detuning as ∣μ∣\sqrt{|\mu|}, for doping values μ\mu close to vH point, the corresponding lengthscales can be in the tens of nanometers (see Figure 1, and also Figure 3 and Figure 4 discussed later). This is much shorter than in typical superconducting metals with a large Fermi surface, falling in the range readily accessible by state-of-the-art STM probes.

TDOS patterns showing vH-selective Friedel oscillations

Figure 3. Probing superconductivity by vH-selective Friedel oscillations. Shown is TDOS pattern on a line-cut through the impurity along the yy direction. Oscillations fall off exponentially, as e−2y/ξe^{-2y/\xi}, with the best-fit lengthscale ξ\xi values quoted in (a), (b) and (c) matching closely the prediction from Eq. (4). Since ξ\xi is proportional to the Fermi velocity, detuning away from the van Hove point (at the same temperature and same gap magnitude) causes slower decay, as demonstrated in panels (a)-(c). From the exponential fits, we reconstruct the gap magnitude using Eq. (17), and find that the agreement with Eq. (4) is better than 15% even at detunings as large as μ=−0.6t\mu=-0.6t, and the disagreement is less than a percent for μ=−0.2t\mu=-0.2t. Panel (d) illustrates that the superconductor has smaller coherence length than the normal metal at a comparable temperature, as expected from Eq. (4). The magnitude of the gap is estimated using Eq. (17), to a high degree of accuracy.

Van Hove-selective Friedel oscillations in square-lattice and triangular-lattice band structures

Figure 4. Van Hove–selective Friedel oscillations arising from van Hove (vH) points in two band structures: (a) a square-lattice t−t′t-t' hopping model with t′=−0.2tt'=-0.2t, and (b) a triangular-lattice nearest-neighbor hopping model. The vH selectivity is evident in both cases: each vH point generates hyperbolic Friedel fringes confined to a distinct pair of real-space sectors [X and Y in (a); M₁, M₂, and M₃ in (b)]. The light cone opening angle is smaller than 90° due to nonzero t′t'. In each case, the hyperbola asymptotes define the sector boundaries, demonstrating that this behavior is generic rather than tied to a particular band-structure model.

Furthermore, as shown below, vH points XX and YY are responsible for light-cone Friedel oscillations in different quadrants, the East-West for XX and North-South for YY. This property provides a convenient diagnostic of ΔX\Delta_X and ΔY\Delta_Y, which can be determined separately and independently. Similarly, for the bandstructures with three

Van Hove-selective Friedel oscillations

Figure 2. Van Hove-selective Friedel oscillations in the presence of superconducting gap functions with different angular dependences: (a) dx2−y2d_{x^2-y^2}, whose maxima occur at the vH points, and (b) dxyd_{xy}, whose maxima lie away from the vH points. The dx2−y2d_{x^2-y^2} gap is maximal at the XX and YY points in kk space, resulting in Friedel oscillation patterns taken at low voltage bias being suppressed along the real-space xx and yy axes. Conversely, the dxyd_{xy} gap is maximal along the momentum-space diagonal directions, suppressing Friedel oscillations along the corresponding real-space diagonals and leaving comparatively stronger oscillations along the xx and yy axes.

vH points, such as those common for bandstructures with hexagonal lattice symmetry (see, e.g., Ref. [9]), light-cone Friedel oscillations from different vH points occur in different 60∘60^\circ sectors, allowing separate diagnostic of the gap function at these vH points.

This spatial separation gives the STM measurement unique diagnostic power. A conventional local tunneling spectrum receives contributions from the entire Fermi surface, so a low-bias suppression alone cannot determine whether the superconducting gap is concentrated near a vH point or elsewhere in momentum space. Here, in contrast, quasiparticles originating near different vH points propagate into distinct, non-overlapping real-space sectors. Comparing the fringe envelope in each sector between the normal and superconducting states therefore provides a momentum-selective measurement without requiring momentum-resolved spectroscopy: suppression in a given sector directly signals a gap for the corresponding vH quasiparticles.

Experimentally, the proposed protocol is therefore as follows. First, a low-bias dI/dVdI/dV map is acquired around an isolated impurity in the normal state, either above TcT_c or with superconductivity suppressed by a magnetic field. The resulting light-cone fringe pattern identifies the real-space sectors associated with each vH point and provides a normal-state reference for the oscillation period and decay length. The same field of view is then measured in the superconducting state at a fixed low bias. By comparing the sector-resolved fringe envelopes to the normal-state reference, one can determine whether superconductivity selectively suppresses quasiparticles originating from a given vH point. Fitting the decay lengths in the normal and superconducting states then yields the corresponding gap magnitude through the relation derived below. Repeating this analysis for the distinct vH sectors allows the gap magnitudes at symmetry-inequivalent saddle points to be determined independently, while the phase dependence between them requires an interferometric measurement involving multiple impurities [65, 66, 67].

Beyond probing superconductivity, the vH-selective Friedel oscillations may provide a real-space probe of anomalous normal-state carrier dynamics near vH points [39, 40, 70]. Interactions near a vH singularity can produce marginal- or non-Fermi-liquid self-energies and anomalously short quasiparticle lifetimes [68, 69]. For a hyperbolic saddle, Houghton et al. found a thermally cut-off equilibrium Friedel response and, separately, a marginal inelastic lifetime near the Lifshitz transition [36]. With thermal and disorder broadening independently calibrated, the temperature- and energy-dependent envelope of the vH-resolved TDOS fringes could test whether anomalous scattering is concentrated near particular vH sectors. The pairing-induced sector suppression studied here then supplies the complementary gap diagnostic, allowing vH-related decoherence and superconducting gap formation to be distinguished in real space. This is an interesting topic for future work.

The analysis of vH Friedel oscillations in a superconductor can be carried out using the BdG Hamiltonian describing quasiparticles near the Fermi level,

HBdG(k)=(Ek−μΔkΔk∗μ−E−k)(5)H_{\mathrm{BdG}}(\mathbf{k}) = \begin{pmatrix} E_{\mathbf{k}}-\mu& \Delta_{\mathbf{k}} \\ \Delta_{\mathbf{k}}^* & \mu-E_{-\mathbf{k}} \end{pmatrix} \tag*{(5)}

The corresponding momentum space Green’s function for the superconductor is,

G(k,iωm)=1iωm−(Ek(k)−μ)σz−Δkσx(6)G(\mathbf{k},i\omega_m)=\frac{1}{i\omega_m-(E_{\mathbf{k}}(\mathbf{k})-\mu)\sigma_z-\Delta_{\mathbf{k}}\sigma_x} \tag*{(6)}

A much-studied superconductivity problem with pairing enhanced by vH points is defined on a 2D square lattice, Eq. (2). The saddle points occur at X=(±π/a,0)X=(\pm\pi/a,0), and Y=(0,±π/a)Y=(0,\pm\pi/a), where aa is the lattice constant. The system is half-filled at μ=0\mu=0, electron-doped at μ>0\mu>0, and hole-doped at μ<0\mu<0. Superconducting gap function centered at vH points is modeled as

ΔkvH=Δ0e−(k−kvH)2/k02.(7)\Delta_{\mathbf{k}}^{\mathrm{vH}}=\Delta_0 e^{-(\mathbf{k}-\mathbf{k}_{\mathrm{vH}})^2/k_0^2}. \tag*{(7)}

This gap function has ss-wave symmetry if ΔX=ΔY\Delta_X=\Delta_Y, dx2−y2d_{x^2-y^2} symmetry if ΔX=−ΔY\Delta_X=-\Delta_Y, or no rotational symmetry (nematic superconducting order) if ΔX≠ΔY\Delta_X\ne\Delta_Y. The results are found to be not particularly sensitive to the gap function peak width parameter k0k_0.

Near the vH point YY, we can write kx=qxk_x=q_x and ky=π+qyk_y=\pi+q_y. Expanding to quadratic order, we obtain the effective dispersion

E(0,π)(q)≈ta2(qx2−qy2)−μ.(8)E_{(0,\pi)}(\mathbf{q})\approx ta^2(q_x^2-q_y^2)-\mu. \tag*{(8)}

Similarly, for the other van Hove point (XX), the effective dispersion is

E(π,0)(q)≈ta2(qy2−qx2)−μ.(9)E_{(\pi,0)}(\mathbf{q})\approx ta^2(q_y^2-q_x^2)-\mu. \tag*{(9)}

Thus, the two saddle points have hyperbolic dispersions with exchanged principal directions.

This simple fact governs the quasiparticle dynamics in real-space. For an ordinary convex Fermi surface, a point source can emit semiclassical quasiparticle trajectories in an approximately isotropic manner. In comparison, near the YY van Hove point, the low-energy quasiparticles predominantly move along yy axis and nearby directions confined to 90°-quadrants. At other spatial directions the semiclassical trajectories are forbidden. The boundaries between the two regions occur at the lines y=±xy=\pm x. The allowed sector is ∣y∣>∣x∣|y|>|x| (‘time-like’ separated region) where the propagator oscillates. Outside this (for ∣x∣>∣y∣|x|>|y|, i.e., the ‘space-like’ separated region) the propagator is suppressed exponentially. For the XX van Hove point, the roles of xx and yy are interchanged: semiclassical trajectories are confined to 90°-quadrants around the xx axis, and are absent outside these quadrants. This is the origin of the light cone effect.

Within the hyperbolic saddle approximation for the bandstructure, the real-space Green’s function due to a van Hove point is

G(0,π)(r)=iωn+ΔσxΩnF0(r)+σzF1(r)(10)G^{(0,\pi)}(\mathbf{r})=\frac{i\omega_n+\Delta\sigma_x}{\Omega_n}F_0(\mathbf{r})+\sigma_zF_1(\mathbf{r}) \tag*{(10)}

where Ωn=ωn2+Δ2\Omega_n=\sqrt{\omega_n^2+\Delta^2}, and F0(r)F_0(\mathbf{r}) and F1(r)F_1(\mathbf{r}) are, respectively, the imaginary and real parts of the modified Bessel function of the second kind

K0(−∣μ∣+iΩnta2y2−x2)(11)K_0\left(\sqrt{-\frac{|\mu|+i\Omega_n}{ta^2}}\sqrt{y^2-x^2}\right) \tag*{(11)}

(sometimes also called the Macdonald function). Similarly, for vH point X=(π,0)X=(\pi,0)

G(π,0)(r)=iωn−ΔσxΩnF0(r)+σzF1(r)(12)G^{(\pi,0)}(\mathbf{r})=\frac{i\omega_n-\Delta\sigma_x}{\Omega_n}F_0(\mathbf{r})+\sigma_zF_1(\mathbf{r}) \tag*{(12)}

For the XX saddle point, the argument in F0F_0 and F1F_1 will be (x2−y2)1/2(x^2-y^2)^{1/2} instead.

Near the YY point, the propagators vanish exponentially in the regions ∣x∣>∣y∣|x|>|y|, a behavior similar to Klein-Gordon relativistic scalar field propagator outside its light cone, and they sinusoidally oscillate in the regions ∣x∣<∣y∣|x|<|y|. This can be understood from the fact that near the van Hove point Y=(0,π)Y=(0,\pi), the semiclassical velocity aligns with the ±y\pm y axis, and the quasiparticle cannot propagate along the xx axis. In a similar manner, near the XX point, the propagators vanish exponentially in the regions ∣x∣<∣y∣|x|<|y| and oscillate in the regions ∣x∣>∣y∣|x|>|y|.

To describe Friedel oscillations near a localized impurity we consider the real-space tunneling density of states (TDOS) dI/dVdI/dV. For conventional (circular) Fermi surfaces, the impurity-modified TDOS features the familiar form of Friedel oscillations consisting of concentric rings. In comparison, the van Hove singularity produces hyperbolic fringes in the Friedel oscillations near an impurity (see Figure 1), of the form expressed in Eq. (3).

The origin of the light-cone Friedel fringes and their spatial structure, can be illustrated by considering impurity-induced density oscillations in the normal state, Δ=0\Delta=0, near a weak point-like impurity U(r)=u0δ(r)U(\mathbf{r})=u_0\delta(\mathbf{r}). To first order in u0u_0, the local Green’s function acquires the correction

δG1R(E;r,r)=u0[G0R(E,r)]2.(13)\delta G_1^{R}(E;\mathbf{r},\mathbf{r})=u_0\left[G_0^{R}(E,\mathbf{r})\right]^2. \tag*{(13)}

For the hyperbolic saddle-point dispersion, the retarded propagator in a semiclassically allowed angular sector is expressed through a Hankel function, as given above, with large-distance asymptotics

G0R(E,r)∝e±i(qEs−π/4)qEs,qE=∣E+μ∣ta2.(14)G_0^{R}(E,\mathbf{r})\propto\frac{e^{\pm i(q_Es-\pi/4)}}{\sqrt{q_Es}},\qquad q_E=\sqrt{\frac{|E+\mu|}{ta^2}}. \tag*{(14)}

Here s=∣x2−y2∣s=\sqrt{|x^2-y^2|} is the Minkowski distance introduced above, restricted to the allowed sector. Integrating the impurity-induced spectral density over occupied energies gives, at zero temperature:

δnosc(R,θ)≃u08π2ta2R2∣cos⁡2θ∣sin⁡(2kFR∣cos⁡2θ∣),(15)\delta n_{\mathrm{osc}}(R,\theta)\simeq\frac{u_0}{8\pi^2ta^2R^2|\cos2\theta|}\sin\left(2k_FR\sqrt{|\cos2\theta|}\right), \tag*{(15)}

where kF=∣μ∣/ta2k_F=\sqrt{|\mu|/ta^2}, x=Rcos⁡θx=R\cos\theta, and y=Rsin⁡θy=R\sin\theta. This result applies for kFs≫1k_Fs\gg1 in the oscillatory sectors of a single saddle-point band, and vanishes in the complementary sectors. The constant-phase contours are hyperbolas, while the amplitude is strongly enhanced toward the light-cone boundaries x=±yx=\pm y. The apparent divergence near these boundaries signals the breakdown of the large-distance asymptotic expansion, a behavior that can be cured by reintroducing the full propagator.

The result above describes the equilibrium density modulation, obtained by integrating the impurity-induced local density of states over occupied energies. In contrast, STM probes the energy-resolved tunneling density of states (TDOS), which at low temperatures is proportional to the differential conductance (dI/dVdI/dV) at the corresponding bias voltage. This distinction affects the spatial decay and phase of the oscillations: at zero temperature, the equilibrium density oscillations decay as 1/s21/s^2, whereas the TDOS oscillations at fixed energy decay as 1/s1/s. Both, however, exhibit the same characteristic hyperbolic geometry. We now calculate the TDOS in the superconducting state using the BdG Green’s functions and the impurity correction in Eq. (13).

Next, we calculate tunneling current in the superconducting state, Δ≠0\Delta\ne0. Using the bare BdG Green’s function given above, Eq. (10), the impurity-modified Green’s function can be computed, and related to TDOS as

dIdV∣V=0=4e2tSTM2Re⁡[Tr⁡∑ωn>0δδG1dz∣z=iωn](16)\left.\frac{dI}{dV}\right|_{V=0}=4e^2t_{\mathrm{STM}}^2\operatorname{Re}\left[\operatorname{Tr}\sum_{\omega_n>0}\left.\frac{\delta\delta G_1}{dz}\right|_{z=i\omega_n}\right] \tag*{(16)}

(see also Ref. [66]). Here we consider point-like impurities which scatter quasiparticles as delta-function potentials, U(r)=u0δ(r−ri)U(\mathbf{r})=u_0\delta(\mathbf{r}-\mathbf{r}_i). In momentum space, each impurity generates a constant interaction that scatters between any two momenta with equal rates.

At half filling, the TDOS signal is found to be concentrated only on the light-cone boundaries. However, slightly away from half-filling, Friedel oscillations take the light-cone form of sharp, hyperbolic fringes. Moreover, since kF∝μk_F\propto\sqrt{\mu}, the distance between consecutive fringes is λF/2\lambda_F/2, which scales as 1/μ1/\sqrt{\mu}.

This is illustrated in Figure 1, obtained by a numerical calculation using the lattice Green’s function of the normal metal phase in a square lattice band. As the hole doping is increased, eventually the Fermi surface becomes nearly circular, described by a conventional pattern of concentric rings.

The observation that the real-space Friedel oscillations due to quasiparticles near vH points are hyperbolic fringes rather than concentric rings, serves as a natural starting point for studying the superconductivity problem. The evolution of these fringes as the system becomes superconducting, can reveal useful information about the superconductive pairing near the vH points.

In particular, when there is strong superconducting pairing near the van Hove points, there is a depletion of quasiparticle density of states at low bias voltage and temperature, and the hyperbolic fringes in the allowed regions (those within “light cones”) are heavily weakened. This is demonstrated in Figure 3, where the Friedel oscillation near an impurity is compared for Δ=0\Delta=0, and Δ=0.0256∣μ∣\Delta=0.0256|\mu|, at a fixed temperature. If the van Hove points host a hotspot for the gap function, then such a suppression of TDOS at certain directions will be observed as the system enters the superconducting phase from the normal metal phase. Such a comparison can help to experimentally determining whether the superconducting gap function is localized near the van Hove points, or away from them. To the contrary, if the gap function is primarily localized away from the van Hove points, then such a strong suppression of TDOS will not be observed in the superconducting state.

A finite temperature and a finite superconducting gap causes exponential decay of the oscillating pattern of Friedel oscillations (see Eq. (3)) of the TDOS at zero bias. The lengthscale of such an exponential decay is governed by the coherence length, Eq. (4) [71, 72, 73]. The gap function’s magnitude can then be recovered by comparing the exponential decay of the Friedel oscillation of TDOS of the normal metal at zero bias and at a temperature T2T_2, to that of the superconductor at zero bias and at a finite temperature T1T_1, using the relation,

Δ2(T1)=(πkBT2)2ξ2(T2,Δ=0)ξ2(T1,Δ(T1))−(πkBT1)2.(17)\Delta^2(T_1)=(\pi k_{\mathrm{B}}T_2)^2\frac{\xi^2(T_2,\Delta=0)}{\xi^2(T_1,\Delta(T_1))}-(\pi k_{\mathrm{B}}T_1)^2 . \tag*{(17)}

Here, the temperature T2T_2 of the metal is usually greater than T1T_1, the temperature of the superconductor. Alternatively, if the superconductivity is suppressed with an external magnetic field, then normal metal is obtained at the same temperature T2T_2.

We note, however, that the relative phases of Δ\Delta at different vH points cannot be recovered with this method. For determining these phases, the quasiparticle interference methods described in Refs.[55, 65, 66, 67], or a similar technique, can be employed.

Lastly, we show that the light-cone Friedel oscillations are not a peculiarity of the nearest-neighbor square-lattice model. Figure 4 compares a square-lattice t−t′t-t' band [ Figure 4(a)] with a nearest-neighbor triangular-lattice band [ Figure 4(b)], whose global Fermi-surface geometries and arrangements of saddle points are qualitatively different. In both cases, however, the response separates into real-space sectors associated with individual vH points.

The square lattice has two inequivalent saddle-point directions, producing pairs of less-than 90° sectors associated with the X and Y points, whereas the three symmetry-related M points of the triangular lattice produce pairs of 60° sectors. As a further test of generality, we checked that the light-cone character of Friedel oscillations in the square lattice remains unchanged for a more general t−t′−t′′t-t'-t'' band model.

The general character of the light-cone behavior and its independence of the band symmetry follows from the generic local structure of a two-dimensional saddle point. After choosing its principal axes, the dispersion has opposite curvatures along the two directions. This hyperbolic geometry confines low-energy quasiparticle propagation to a pair of real-space sectors, while propagation in the complementary directions is exponentially suppressed. Crystal symmetry determines how the sectors associated with the different vH points are oriented and tiled in real space, but without constraining the sectors’ selectivity. Thus, even when the global band structure differs substantially from the idealized square-lattice case, an STM map can identify the spatial sectors that selectively probe quasiparticles near each vH point.

In conclusion, near an impurity, each van Hove point maps onto its own real-space “light-cone” sector, allowing STM around a single impurity to measure whether—and how strongly—that saddle-point quasiparticle sector is gapped by superconducting order. Namely, quasiparticles near vH points generate distinctive real-space signatures in the tunneling density of states near impurities. The unusual momentum-space geometry of the van Hove points is reflected in the strongly anisotropic real-space propagation in both the normal metal and the superconductor. The hyperbolic dispersion near the van Hove points restricts the possible quasiparticle velocity direction, which effectively divides real space into sectors where propagation is allowed, resulting in oscillating Green’s function, and forbidden sectors where propagators exponentially decay over a short length scales. The resulting Friedel oscillations of TDOS near an impurity display hyperbolic patterns rather than conventional maps consisting of concentric circles. Strong superconducting pairing near the van Hove points suppress the hyperbolic fringes in Friedel oscillations as no quasiparticles modes near van Hove points are available at a low-energy. In comparison, when the gap function is concentrated away from the van Hove regions, the characteristic hyperbolas in Friedel oscillations remain very similar to its normal metal counterpart. In this manner, the single-impurity interference can be utilized to experimentally distinguish whether the gap function is strongly localized at van Hove points, or away from them.

We acknowledge useful discussions with Andrey Chubukov, Leonid Glazman, Peter Hirschfeld, Patrick Lee and Boris Spivak.

References

References

  1. [1]L. Van Hove, The Occurrence of Singularities in the Elastic Frequency Distribution of a Crystal, Phys. Rev. 89, 1189 (1953).DOI
  2. [2]R. S. Markiewicz, A survey of the van Hove scenario for high-Tc superconductivity with special emphasis on pseudogaps and striped phases, J. Phys. Chem. Solids 58, 1179 (1997).
  3. [3]K. G. F. Ron, J. C. Campuzano, A. A. Abrikosov, M. Lindroos, A. Bansil, H. Ding, D. Koelling, and B. Dabrowski, Observation of an “extended” van Hove singularity in YBa2Cu4O8 by ultrahigh energy resolution angle-resolved photoemission, Phys. Rev. Lett. 73, 3302 (1994).
  4. [4]J. González, F. Guinea, and M. A. H. Vozmediano, Marginal-Fermi-liquid behavior from two-dimensional Coulomb interaction, Phys. Rev. B 59, R2474-R2477 (1999).DOI
  5. [5]A. Piriou, N. Jenkins, C. Berthod, I. Maggio-Aprile, and Ø. Fischer, First direct observation of the van Hove singularity in the tunnelling spectra of cuprates, Nat. Commun. 2, 221 (2011).DOI
  6. [6]G. Levy, C. Berthod, A. Piriou, E. Giannini, and Ø. Fischer, Preeminent role of the van Hove singularity in the strong-coupling analysis of scanning tunneling spectroscopy for two-dimensional cuprate superconductors, Phys. Rev. Lett. 101, 267004 (2008).pubmed.ncbi.nlm.nih.gov/19437664
  7. [7]D. M. Newns, C. C. Tsuei, R. P. Hueberner, P. J. M. van Bentum, P. C. Pattnaik, and C. C. Chi, Quasiclassical Transport at a van Hove Singularity in Cuprate Superconductors, Phys. Rev. Lett. 73, 1695 (1994).
  8. [8]A. Steppke, L. Zhao, M. E. Barber, T. Scaffidi, F. Jermembeck, H. Rosner, A. S. Gibbs, Y. Maeno, S. H. Simon, A. P. Mackenzie, and C. W. Hicks, Strong peak in Tc of Sr2RuO4 under uniaxial pressure, Science 355, eaaf9398 (2017).
  9. [9]R. Nandkishore, L. S. Levitov, and A. V. Chubukov, Chiral superconductivity from repulsive interactions in doped graphene, Nat. Phys. 8, 158 (2012).DOI
  10. [10]R. Comin, A. Frano, M. M. Yee, Y. Yoshida, H. Eisaki, E. Schierle, E. Weschke, R. Sutarto, F. He, A. Soumyanarayanan, Y. He, M. Le Tacon, I. S. Elfimov, J. E. Hoffman, G. A. Sawatzky, B. Keimer, and A. Damascelli, Charge Order Driven by Fermi-Arc Instability in Bi2Sr2−xLaxCuO6+δ, Science 343, 390–392 (2014).
  11. [11]M. H. Hamidian, S. D. Edkins, S. H. Joo, A. Kostin, H. Eisaki, S. Uchida, M. J. Lawler, E.-A. Kim, A. P. Mackenzie, K. Fujita, J. Lee, and J. C. S. Davis, Detection of a Cooper-pair density wave in Bi2Sr2CaCu2O8+x, Nature 532, 343–347 (2016).DOI
  12. [12]S. D. Edkins, A. Kostin, K. Fujita, A. P. Mackenzie, H. Eisaki, S. Uchida, S. Sachdev, M. J. Lawler, E.-A. Kim, J. C. Séamus Davis, and M. H. Hamidian, Magnetic field–induced pair density wave state in the cuprate vortex halo, Science 364, 976–980 (2019).
  13. [13]T. Timusk and B. Statt, The pseudogap in high-temperature superconductors: an experimental survey, Rep. Prog. Phys. 62, 61 (1999).DOI
  14. [14]M. R. Norman, H. Ding, M. Randeria, J. C. Campuzano, T. Yokoya, T. Takeuchi, T. Takahashi, T. Mochiku, K. Kadowaki, P. Guptasarma, and D. G. Hinks, Destruction of the Fermi surface in underdoped high-Tc superconductors, Nature 392, 157–160 (1998).DOI
  15. [15]A. Kanigel, M. R. Norman, M. Randeria, U. Chatterjee, S. Souma, A. Kaminski, H. M. Fretwell, S. Rosenkranz, M. Shi, T. Sato, T. Takahashi, Z. Z. Li, H. Raffy, K. Kadowaki, D. Hinks, L. Ozyuzer, and J. C. Campuzano, Evolution of the pseudogap from Fermi arcs to the nodal liquid, Nat. Phys. 2, 447–451 (2006).DOI
  16. [16]A. Damascelli, Z. Hussain, and Z.-X. Shen, Angle-resolved photoemission studies of the cuprate superconductors, Rev. Mod. Phys. 75, 473 (2003).DOI
  17. [17]B. Keimer, S. A. Kivelson, M. R. Norman, S. Uchida, and J. Zaanen, From quantum matter to high-temperature superconductivity in copper oxides, Nature 518, 179 (2015).DOI
  18. [18]N. Furukawa, T. M. Rice, and M. Salmhofer, Truncation of a two-dimensional Fermi surface due to quasiparticle gap formation at the saddle points, Phys. Rev. Lett. 81, 3195 (1998).DOI
  19. [19]K. Le Hur and T. M. Rice, Superconductivity close to the Mott state: From condensed-matter systems to superfluidity in optical lattices, Ann. Phys. (NY) 324, 1452 (2009).DOI
  20. [20]H. J. Schulz, Superconductivity and antiferromagnetism in the two-dimensional Hubbard model: scaling theory, Europhys. Lett. 4, 609 (1987).DOI
  21. [21]J. E. Hirsch and D. J. Scalapino, Enhanced Superconductivity in quasi Two-Dimensional Systems, Phys. Rev. Lett. 56, 2732 (1986).pubmed.ncbi.nlm.nih.gov/10033076
  22. [22]I.E. Dzyaloshinskii, Maximal increase of the superconducting transition temperature due to the presence of van Hove singularities, JETP Lett., 46 (3), 118-121 (1987).
  23. [23]I.E. Dzyaloshinskii, Superconductivity of compounds based on La2CuO4, JETP Lett., 46(S1), S93-S97 (1987)
  24. [24]I. E. Dzyaloshinskii, Maximal increase of the superconducting transition temperature due to the presence of van Hove singularities, Sov. Phys. JETP 66, 848 (1987).
  25. [25]J. Gonz’alez and T. Stauber, “Marginal Fermi liquid in twisted bilayer graphene,” Phys. Rev. Lett. 124, 186801 (2020).DOI
  26. [26]A. Chronister, M. Zingl, A. Pustogow, Y. Luo, D. A. Sokolov, F. Jerzembec, N. Kikugawa, C. W. Hicks, J. Mavrlje, E. D. Bauer, J. D. Thompson, A. P. Mackenzie, A. Georges, and S. E. Brown, Tuning the Fermi liquid crossover in Sr2RuO4 with uniaxial stress, npj Quantum Mater. 7, 113 (2022).arxiv.org/abs/2111.05570
  27. [27]A. Hunter, C. Putzke, F. B. Kugler, S. Beck, E. Cappelli, F. Margot, M. Straub, Y. Alexanian, J. Teyssier, A. de la Torre, K. W. Plumb, M. D. Watson, T. K. Kim, C. Cacho, N. C. Plumb, M. Shi, M. Ravici, J. Osiecki, C. Polley, D. A. Sokolov, A. P. Mackenzie, E. Berg, A. Georges, P. J. W. Moll, A. Tamai, and F. Baumberger, Non-Fermi liquid quasiparticles in strain-tuned Sr2RuO4, arXiv:2503.11311 (2025).arxiv.org/abs/2503.11311
  28. [30]M. I. Salkola, A. V. Balatsky, and D. J. Scalapino, Theory of Scanning Tunneling Microscopy Probe of Impurity States in a d-Wave Superconductor, Phys. Rev. Lett. 77, 1841 (1996).DOI
  29. [31]A. V. Balatsky, I. Vekhter, and J.-X. Zhu, Impurity- induced states in conventional and unconventional su- perconductors, Rev. Mod. Phys. 78, 373 (2006).DOI
  30. [32]G. D. Mahan, Friedel oscillations on a square lattice, Int. J. Mod. Phys. B 9, 1327–1341 (1995).DOI
  31. [33]E. C. Andrade, E. Miranda, and V. Dobrosavljević, Quantum ripples in strongly correlated metals, Phys. Rev. Lett. 104, 236401 (2010).pubmed.ncbi.nlm.nih.gov/20867254
  32. [34]A. Weismann, M. Wenderoth, S. Lounis, P. Zahn, N. Quaas, R. G. Ulbrich, P. H. Dederichs, and S. Bl”ugel, Seeing the Fermi surface in real space by nanoscale elec- tron focusing, Science 323, 1190–1193 (2009).DOI
  33. [35]S. Lounis, P. Zahn, A. Weismann, M. Wenderoth, R. G. Ulbrich, I. Mertig, P. H. Dederichs, and S. Bl”ugel, Theory of real space imaging of Fermi surface parts, Phys. Rev. B 83, 035427 (2011).DOI
  34. [36]C. C. A. Houghton, E. G. Mishchenko, and M. E. Raikh, Enhanced interaction effects in the vicinity of the topo- logical transition, Phys. Rev. B 100, 165140 (2019).DOI
  35. [37]C. Bena, Friedel oscillations: Decoding the hidden physics, Comptes Rendus Physique 17, 302–321 (2016).
  36. [38]C.-K. Lu, Friedel oscillation near a van Hove singularity in two-dimensional Dirac materials, J. Phys.: Condens. Matter 28, 065001 (2016).
  37. [39]L. Wei, Q. Xu, Y. He, Q. Li, Y. Huang, W. Zhu, K. Watanabe, T. Taniguchi, M. Claassen, D. A. Rhodes, D. M. Kennes, L. Xian, A. Rubio, and L. Wang, Lin- ear resistivity at van Hove singularities in twisted bilayer WSe₂, Proc. Natl. Acad. Sci. U.S.A. 121, e2321665121 (2024).DOI
  38. [40]M. E. Barber, A. S. Gibbs, Y. Maeno, A. P. Mackenzie, and C. W. Hicks, Resistivity in the vicinity of a van Hove singularity: Sr₂RuO₄ under uniaxial pressure, Phys. Rev. Lett. 120, 0776602 (2018).
  39. [55]P. J. Hirschfeld, D. Altenfeld, I. Eremin, and I. I. Mazin, Robust determination of the superconducting gap sign structure via quasiparticle interference, Phys. Rev. B 92, 184513 (2015).DOI
  40. [65]P. Ding, T. Schwemmer, C. H. Lee, X. Wu, and R. Thomale, Hyperbolic Fringe Signal for Twin Impurity Quasiparticle Interference, Phys. Rev. Lett. 130, 256001 (2023).DOI
  41. [66]A. Panigrahi, V. Poliakov, and L. Levitov, Tomographic imaging of superconducting order using particle–hole interference, Proc. Natl. Acad. Sci. U.S.A. 123 (27), e2534730123 (2026).DOI
  42. [67]A. Panigrahi, V. Poliakov, and L. Levitov, Particle-Hole Ghost Interference in Superconductors, arXiv preprint arXiv:2606.06437 (2026).arxiv.org/abs/2606.06437
  43. [68]A. Kapustin, T. McKinney, and I. Z. Rothstein, “Wilsonian effective field theory of two-dimensional Van Hove singularities,” Phys. Rev. B 98, 035122 (2018).DOI
  44. [69]A. A. Katanin and A. P. Kampf, “Quasiparticle anisotropy and pseudogap formation from the weak-coupling renormalization group point of view,” Phys. Rev. Lett. 93, 106406 (2004).pubmed.ncbi.nlm.nih.gov/15447431
  45. [70]A. Jaoui, I. Das, G. Di Battista, J. Diez-Merida, X. Lu, K. Watanabe, T. Taniguchi, H. Ishizuka, L. Levitov, and D. K. Efetov, “Quantum critical behaviour in magic-angle twisted bilayer graphene,” Nat. Phys. 18, 633–638 (2022).DOI
  46. [71]J. M. Byers, M. E. Flatté, and D. J. Scalapino, Influence of gap extrema on the tunneling conductance near an impurity in an anisotropic superconductor, Phys. Rev. Lett. 71, 3363 (1993).DOI
  47. [72]T. Das, J.-X. Zhu, and M. J. Graf, Local suppression of the superfluid density of PuCoGa₅ by strong onsite disorder, Phys. Rev. B 84, 134510 (2011).DOI
  48. [73]L. Zhu, W. A. Atkinson, and P. J. Hirschfeld, Two impurities in ad-wave superconductor: Local density of states, Phys. Rev. B 67, 094508 (2003).DOI

Paper details

vlnc:W_yemxtfm3
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 11

downloads · 2

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