We present an exact spectral and dynamical analysis of continuous-time quantum walks on barbell graphs and two variants of star-of-cliques graphs, obtaining closed-form eigenvalues, eigenvectors, eigenstate inverse participation ratios (IPRs), and long-time dynamical IPRs ($\overline{\mathrm{IPR}}$) for all families and vertex types. The central finding is that localization in these graphs arises through three distinct mechanisms, each responsible for a different vertex class: null-coupling confinement, in which dark modes have exactly zero amplitude at the intersubgraph connection and are therefore confined within individual cliques for all $n$; destructive-interference standing waves, whose opposite phase on either side of a structural bottleneck suppresses amplitude flow across it; and invariant-subspace dynamical localization, in which a large degenerate eigenspace drives $\overline{\mathrm{IPR}}\to 1$ for walks initialized at clique interior vertices—a result we establish through exact, vertex-specific dynamical IPR, in contrast to the graph-averaged treatments of degeneracy-driven trapping in prior work. In the barbell, clique interior vertices achieve $\overline{\mathrm{IPR}}\to 1$ while bridge vertices saturate at $1/2$. In variant 1 of the star-of-cliques construction, where every clique vertex connects to the hub, both hub and clique vertices achieve $\overline{\mathrm{IPR}}\to 1$. Restricting each clique to a single bridge connection (variant 2) reduces the hub's $\overline{\mathrm{IPR}}$ from 1 to $1/4$—fixed by its participation in only three eigenmodes regardless of graph size—while bridge and clique vertices retain $\overline{\mathrm{IPR}}\to 1$. Each mechanism responds differently to structural perturbation: The null-coupling mechanism is exact; the interference mechanism relies on the reflection symmetry between the two sides of the bottleneck; and the invariant-subspace mechanism is asymptotic in graph size. These results provide analytically solvable examples of how spectral degeneracy and connectivity together determine long-time quantum transport in modular networks.

Close

PrevNext

View figure in article

• Quantum algorithms & computation
• Quantum computation
• Quantum information theory