A spectral viewpoint on the single defect tight-binding chain

Abstract

We analyze the time evolution of the nearest-neighbour tight-binding chain in the presence of a single onsite defect. Such a defect was shown to generate non-trivial transport behavior recently in the article \textit{Acharya et al J. Stat. Mech. (2026) 043102}. The authors have used a defect technique inspired by classical random walk methods to obtain exact analytical expressions for the occupation probability and subsequently the mean and mean-squared displacement (MSD). Here we derive the same results using a spectral decomposition approach. Starting from the secular equation, we obtain the self-consistency condition for the eigenvalues and construct the corresponding normalized eigenvectors. This approach naturally separates the Hilbert space into dark subspace, whose states have zero amplitude at the defect site and remain unaffected, and bright subspaces, whose states get modified because of the defect. Using this eigenvalue decomposition, we provide a spectral origin of non-monotonicity in the MSD. Numerical calculations for finite chains show that the analytically estimated critical defect strength is in excellent agreement with the defect strength that minimizes the MSD.