Computational Study of the Spectral Properties of Isospectrally Patterned Lattices (2507.08351v1)

Abstract: We perform a computational spectral analysis of different isospectrally patterned lattices (IPL). Having been introduced very recently, the lattice Hamiltonian of IPL consist of coupled cells which possess all the same set of eigenvalues. The latter is achieved in a controllable manner by parametrizing the cells via the phases of the involved orthogonal (or unitary) transformations. This opens the doorway of systematically designing lattice Hamiltonians with unique properties by choosing correspondingly varying phases across the lattice. Here we focus on two-dimensional cells and explore symmetric as well as asymmetric IPL w.r.t. a given center. A tunable fraction of localized vs. delocalized eigenstates belonging to the three subdomains of the corresponding energy bands is demonstrated and analyzed with different measures of localization. In the asymmetric case the center of localization can be shifted arbitrarily by shifting the underlying phase grid. Introducing a complete phase revolution leads for low and high energies to two well-separated branches of localized states which finally merge with increasing energy into the branch of delocalized states. Remarkably, the localized states appear in near-degenerate pairs and this near-degeneracy is lifted upon entering the delocalization regime. A corresponding generalization to several phase revolutions is provided showing a characteristic nodal pattern among the near-degenerate eigenstates.