---
title: 'Tilted Dirac Cone: Theory, Geometry & Applications'
url: https://www.emergentmind.com/topics/tilted-dirac-cone
type: topic
---

# Tilted Dirac Cone: Theory, Geometry & Applications

A tilted Dirac cone is a linear band crossing whose low-energy dispersion contains, in addition to the usual Dirac term, a momentum-linear contribution proportional to the identity matrix, so that locally
\[
\varepsilon(\mathbf q)\sim \mathbf t\cdot\mathbf q \pm \sqrt{v_x^2 q_x^2+v_y^2 q_y^2}.
\]
The crossing therefore remains gapless, but the cone axis is no longer vertical in energy–momentum space. This seemingly simple deformation changes constant-energy contours, Landau quantization, symmetry structure, optical selection, tunneling kinematics, and the topology of the Fermi surface, and it appears in condensed-matter, photonic, circuit, and holographic settings alike [2501.10434][1201.5175][2110.01906].

## 1. Definition, kinematics, and geometric classification

Tilt should be distinguished from anisotropy. Anisotropy means that the Dirac velocities differ by direction, as in \(v_x\neq v_y\), whereas tilt is the identity-matrix term linear in momentum. In the language used across the literature, the same physical quantity is denoted by symbols such as \(\eta\), \(\zeta\), \(\tau\), or \(t\), but the operative meaning is the same: both bands are shifted in the same momentum direction without opening a gap [2411.19226][2112.00542].

The geometric consequence is most transparent in constant-energy cuts. For an upright cone they are circles or ellipses, depending on anisotropy. For a tilted cone they are displaced and, as the tilt grows, pass through a sequence of conic sections. In the barrier formulation
\[
q_x^2+\left(1-\tau^2\right)k_y^2+2\tau k_y(\epsilon-V)-(\epsilon-V)^2=0,
\]
the contour is an ellipse for \(0<\tau<1\), a parabola for \(\tau=1\), and a hyperbola for \(\tau>1\) [2404.02485].

| Regime | Condition | Constant-energy contour |
|---|---:|---|
| Upright | \(\zeta=0\) | Circle or ellipse |
| Type-I / subcritical | \(0<|\zeta|<1\) | Closed ellipse |
| Type-III / critical | \(|\zeta|=1\) | Parabola |
| Type-II / overtilted | \(|\zeta|>1\) | Hyperbola / open Fermi surface |

This classification is not merely descriptive. In type-I cones the Fermi surface at the Dirac point remains pointlike. In type-II cones the Fermi surface is open, and large momentum-space regions can be Pauli blocked even when the chemical potential is tuned to the level crossing point [2112.00542]. In photonic orbital graphene, the critical type-III case appears as a cone with flat dispersion in one direction and linear dispersion in the other, making it the transition point between type-I and type-II behavior [1807.08650].

## 2. Effective Hamiltonians and the origin of tilt terms

For a conventional two-dimensional massless Dirac fermion in a magnetic field,
\[
H_0=v_F(\sigma_x\pi_x+\sigma_y\pi_y),\qquad \bm\pi=\bm p+e\bm A,
\]
with \([\pi_x,\pi_y]=-ie\hbar B\). A minimal tilted generalization is
\[
H(\eta)=-\eta v_F\sigma_0\pi_x+v_F(\sigma_x\pi_x+\sigma_y\pi_y),
\]
where the new term is proportional to \(\sigma_0\) and therefore tilts rather than gaps the cone [1201.5175].

This identity-matrix structure also clarifies a common source of confusion: a gap and a tilt are independent deformations. In the superconducting honeycomb circuit lattice, the parent nearest-neighbor coupling generates the upright cone, while a same-sublattice, axially preferred second-neighbor coupling contributes only a term proportional to \(\sigma_0\), i.e. a tilt rather than a \(\sigma_z\) mass term. In that platform the designed tilt obeys
\[
\zeta=\frac{2J'}{J}=\frac{2L}{L'},
\]
so the tilt is set directly by the ratio of first- and second-neighbor couplings or inductances [2501.10434].

The same separation between gap and tilt is explicit in gate-defined graphene superlattices. There the low-energy satellite cones are described by
\[
\mathcal H'(q)=s_K s\frac{E_g}{2}\sigma_z+s\hbar v\big(t q_y\mathbb I+q_y\sigma_y+s_K T q_x\sigma_x\big),
\]
where the identity term \(s\hbar v\,t q_y\mathbb I\) is the tilt term, the \(\sigma_z\) term is the gap, and the \(T\) coefficient measures anisotropy. The tilt is controlled by the unequal slopes of barrier and well guided modes,
\[
t=\frac{v_w+v_b}{2v},\qquad v=\frac{v_w-v_b}{2},
\]
while the gap is produced by broken reflection symmetry of the superlattice dimerization [2411.19226].

A one-dimensional analogue also exists. In sodium-terminated zigzag graphene nanoribbons at half coverage, the low-energy crossing is fitted by
\[
E_{\lambda}(k)=w_0(k-k_0)+\lambda\,w\,|k-k_0|,
\]
with \(w_0=-2.277\ \mathrm{eV}\times\mathrm{\AA}\), \(w=2.517\ \mathrm{eV}\times\mathrm{\AA}\), and \(k_0=0.469\ \mathrm{\AA}^{-1}\). The nonzero \(w_0\) is the one-dimensional tilt parameter, and the paper emphasizes an asymmetry of about a factor of \(20\) between the branch slopes [1105.0547].

## 3. Symmetry, ellipticity, and topological structure

For upright cones, ordinary chiral symmetry is generated by the Hermitian involution \(\Gamma=\sigma_z\), which anticommutes with \(H_0\). Tilt breaks this relation because the \(\sigma_0\) term cannot anticommute with \(\sigma_z\). For the tilted Hamiltonian above, the appropriate replacement is the non-Hermitian generalized chiral operator
\[
\gamma=\lambda_\eta^{-1}(\sigma_z-i\eta\sigma_y),\qquad \lambda_\eta=\sqrt{1-\eta^2},
\]
which satisfies
\[
\gamma^2=1,\qquad \gamma^\dagger H(\eta)\gamma=-H(\eta)\qquad (|\eta|<1).
\]
This generalized chiral symmetry protects the zero modes and the \(n=0\) Landau level even in the presence of gauge disorder, and it reduces the zero-energy problem to first-order Aharonov–Casher equations [1201.5175].

The same work makes the restriction \(|\eta|<1\) mathematically precise by relating it to ellipticity. The principal symbol
\[
\Xi(\xi_x,\xi_y)=
\begin{pmatrix}
-\eta \xi_x & \xi_x-i\xi_y\\
\xi_x+i\xi_y & -\eta \xi_x
\end{pmatrix}
\]
has determinant
\[
\det\Xi=-(1-\eta^2)\xi_x^2-\xi_y^2.
\]
Hence the tilted Dirac operator is elliptic iff \(|\eta|<1\), exactly the regime in which \(\lambda_\eta\) is real and the generalized chiral construction exists. This is the explicit bridge from tilted-cone zero modes to index-theorem reasoning [1201.5175].

Tilt also coexists with nontrivial topological charges. In photonic orbital graphene, a quadratic touching between a flat band and a parabolic band at zero strain carries winding number \(W=2\); under strain it splits into two tilted Dirac cones, each with \(W=1\). In the same system, further tuning produces type-III cones at the critical tilt [1807.08650]. In a holographic realization, the retarded Green’s function depends on the shifted combination \(\omega+\zeta_x k_x\), giving the inferred dispersion
\[
\omega=-\zeta_x k_x\pm |\mathbf k|,
\]
yet the Berry curvature and Chern number remain
\[
\mathcal F_i=-\frac{k_i}{2|\mathbf k|^3},\qquad C_n=-1,
\]
independent of the tilt parameter [2509.03033].

## 4. Microscopic and synthetic realizations

Tilted Dirac cones can arise from chemistry, lattice geometry, orbital structure, external gating, or synthetic couplings. The mechanisms are diverse, but they repeatedly reduce to one of two low-energy routes: an identity-matrix term produced directly by symmetry-allowed same-sublattice coupling, or an effective identity term generated after downfolding higher-energy degrees of freedom.

| Platform | Mechanism of tilt | Representative feature |
|---|---|---|
| \(8Pmmn\) borophene | Ridge-mediated effective second-neighbor hoppings \(\tilde t-\bar t\) on an emergent honeycomb graph [2108.08183] | \(\zeta_y=\pm 2(\tilde t-\bar t)/t\) |
| Na-terminated \(9\)-ZGNR | Na adsorption at half coverage on alternating bridge sites [1105.0547] | Single Dirac-like crossing at \(k_0=0.469\ \mathrm{\AA}^{-1}\) |
| Superconducting circuit honeycomb lattice | Uniaxial second-neighbor coupling in a \(731\)-site LC resonator array [2501.10434] | Tilt values up to \(59\%\) of relative opposite-direction velocity difference |
| Lieb–kagome photonic lattice | Continuous lattice shear \(\theta\) creating inequivalent \(A\)-\(C\) hoppings [2211.04548] | Upper cones generally type-I, lower cones generally type-II |
| Photonic orbital graphene | Strain-induced splitting of flat-parabolic touching in \(p_x,p_y\) honeycomb bands [1807.08650] | Tilted and type-III cones from a \(W=2\) parent touching |
| Lateral graphene superlattice | Planar array of bipolar waveguides with barrier and well guided modes [2411.19226] | Gate-tunable gapped and tilted satellite cones in the THz range |

The borophene case is unusually microscopic. The low-energy states live on an emergent honeycomb lattice of inner-site \(p_z\) orbitals, while ridge-site \(p_x\) and \(p_z\) orbitals are integrated out. In the resulting two-band model, effective third-neighbor hoppings shift the Dirac point and renormalize anisotropic velocities, while effective second-neighbor hoppings generate the tilt through \(\tilde t-\bar t\). Carbon substitution on ridge sites tends to increase the tilt, while substitution on inner sites tends to decrease it [2108.08183].

The photonic Lieb–kagome and orbital-graphene systems emphasize a complementary route: tilt emerges from geometric deformations that move Dirac points in momentum space. In the sheared Lieb–kagome lattice, the motion of each Dirac point away from the parent high-symmetry point is identified as the origin of the tilt, and the diffraction signature becomes asymmetric and transversely shifted when only cones tilted in one direction are excited [2211.04548]. In orbital graphene, the striking feature is that the tilted cones arise from the splitting of a flat-parabolic touching rather than from a crossing of two ordinary linear bands [1807.08650].

## 5. Transport, optical response, and spectroscopic signatures

Tilt leaves a broad imprint on response functions. In the organic conductor \(\alpha\)-\((\mathrm{BEDT\!-\!TTF})_2\mathrm I_3\), the dynamical polarization function acquires cusps and nonmonotonic structures, and depends not only on the magnitude but also the direction of the external momentum \(\mathbf q\). The same tilted-cone geometry modifies the optical conductivity, plasma frequency, and Coulomb screening, in direct contrast to isotropic graphene [1004.3096].

Optical absorption is especially diagnostic because the interband matrix element and the Pauli-blocked phase space respond differently to tilt. For a generic tilted, anisotropic cone with Hamiltonian
\[
H(\mathbf q)=\hbar v_F(\gamma\eta q_x\sigma_0+\eta q_x\sigma_x+q_y\sigma_y),
\]
the absorption of linearly polarized light depends on both the tilt \(\gamma\) and the velocity anisotropy \(\eta\). Type-II cones retain strong polarization and frequency dependence even at high frequency because open Fermi surfaces leave many states Pauli blocked; type-I cones do not. On that basis an “optical recipe” was proposed to extract both the tilt and the Fermi-velocity anisotropy solely from the absorption spectrum [2112.00542].

In electron optics, tilt reshapes the overlap of propagating states across interfaces. For a \(y\)-tilted barrier in a graphene–tilted-Dirac-material–graphene heterostructure,
\[
H=\left(-i\partial_x\,\sigma_x+\kappa_y\,\sigma_y\right)+\left(\tau \kappa_y+U(x)\right)I_2,
\]
the barrier Fermi surface changes from a circle to an ellipse, parabola, or hyperbola as \(\tau\) crosses \(0\), \(1\), and \(>1\). The overlap with the lead Fermi surface defines an “active surface,” and the deformation of that overlap produces beam collimation, directional asymmetry, and an additional perfect-transmission line at
\[
k_y=-\frac{V}{\tau},
\]
besides the usual normal-incidence Klein channel [2404.02485].

Abrupt changes of tilt also affect pseudospin. In a heterojunction between upright and tilted cones with
\[
\mathcal H=v_x k_x \sigma_x+v_y k_y \sigma_y+v_t k_x \mathbf 1,
\]
normal-incidence Klein tunneling survives in the subcritical regime, but oblique transmission becomes strongly angle and direction dependent, and a tilt discontinuity rotates the pseudospin through a nontrivial matching matrix \(M\) [2312.11262]. In bipolar electron waveguides in \(8\)-\(Pmmn\) borophene, the same tilt allows both barriers and wells to confine states; the overlap of these mode families produces tunable pseudogaps and valley-sensitive guided spectra [2402.08600]. In symmetric double-barrier structures with two tilted regions, coupling between the barriers and the central well gives multiple resonance peaks, including line-type resonances even inside nominally forbidden energy zones [2508.17220].

## 6. Disorder, interactions, and emergent-geometry perspectives

Tilt changes the renormalization-group fate of disorder. In a two-dimensional type-I tilted Dirac semimetal, random scalar potential and random vector potential along the tilting direction cannot exist on their own: they dynamically generate a new disorder that dominates at low energies, turns the system into a compressible diffusive metal, gives the fermions a finite disorder scattering rate, and replaces the isolated band-touching point by a bulk Fermi arc. By contrast, random mass and random vector potential along the non-tilting direction can exist individually, suppress tilt at low energies, and do not produce a bulk Fermi arc. With Coulomb interaction included, those latter cases flow to a stable quantum critical state with finite anomalous dimension and dynamical exponent \(z=1\) [1807.06536].

At charge neutrality, the Boltzmann kinetic theory of interacting tilted Dirac fermions predicts that the Drude pole broadening is enhanced by
\[
\kappa(\zeta)\,(1-\zeta^2)^{-1/2},
\]
while the Drude intensity is anisotropically enhanced by
\[
(1-\zeta^2)^{-1}.
\]
The paper interprets the ubiquitous factor \((1-\zeta^2)^{1/2}\) as a redshift factor of an emergent tilted-material spacetime, and the additional \(\kappa(\zeta)\) as a genuinely interaction-induced enhancement [2205.14175].

That geometric interpretation is developed further in circuit and holographic settings. In the LC-circuit realization, the long-wavelength line element is written as
\[
ds^2=-v_F^2dt^2+\bigl(d\mathbf r-\boldsymbol\zeta v_F dt\bigr)^2,
\]
so the tilt maps directly to off-diagonal metric components \(g_{0i}\), and spatially varying inductances or capacitances produce a spatially varying effective metric [2110.01906]. In holographic hydrodynamics of a tilted Dirac fluid, the same boundary metric leads to anisotropic stress tensor components, reduced shear viscosity, and
\[
\frac{\eta}{s}=\frac{\hbar}{4\pi k_B}(1-\zeta^2)^{3/2},
\]
which is below the usual KSS value for any nonzero tilt [2211.15289]. A related holographic fermion construction finds that optical conductivity develops a pronounced Drude-like peak in the type-II regime even at zero chemical potential, while the Chern number remains tilt independent [2509.03033].

Several limitations recur across this literature. Many symmetry, index-theorem, and kinetic-theory constructions are explicitly restricted to the subcritical regime \(|\eta|<1\) or \(0\le \zeta<1\), where constant-energy contours remain closed and the Dirac operator is elliptic [1201.5175][2205.14175]. Overtilting generally requires separate treatment because it changes Fermi-surface topology, invalidates some zero-mode arguments, and can qualitatively alter transport, optical selection, and collective response. A further recurring point is that tilt should not be conflated with a mass gap: the former is an identity-matrix deformation, whereas the latter is a \(\sigma_z\)-type deformation, and the two can coexist but are distinct control parameters [2411.19226][2501.10434].

Source: https://www.emergentmind.com/topics/tilted-dirac-cone