Engineering Dirac interface states

Abstract

We develop a low-energy theory of interface states in anisotropic multivalley Dirac systems whose masses and kinetic parameters are allowed to vary across an interface. For sharp interfaces, current-conserving matching conditions yield analytic expressions for the existence, localization and dispersion of the bound states. We show that the interface velocity is determined by the weighted tangential kinetic terms on the two sides of the seam. Their cancellation can suppress the linear velocity and generate an interface band that is flat to leading order near the projected Dirac point. For the special antisymmetric configuration in which both the Dirac mass and the tangential kinetic coefficient reverse sign with unchanged magnitude, the transparent sharp-interface solution is exactly dispersionless for all conserved momenta within the linear Dirac theory, while the surrounding bulk bands remain dispersive. We extend the theory to smooth interfaces, where the modified bound-state envelope changes the linear interface velocity through a spatial average of the tangential kinetic coefficient. We also investigate quadratic corrections in the kinetic \(σ_x\) and \(σ_y\) channels. To first order in their coefficients and through linear order in the interface momentum, these terms shift the interface-state energy but produce no additional correction to the linear velocity. Finally, we combine continuum and lattice models to show how interface modes from distinct valleys hybridize and how the resulting dispersions depend on microscopic interface properties. Our results establish design principles for controlling the dispersion, localization, and hybridization of Dirac interface states. We further examine two graphene-based mass-domain-wall models as experimentally inspired examples of dispersive copropagating and counterpropagating interface states.