A particle jumps between the nodes of a graph interacting with local spins. We show that the entanglement entropy of the particle with the spin network is related to the length of the minimum cycle basis. The structure of the thermal state is reminiscent of the string net of spin liquids.
Motivated by recent advances in quantum technologies, we investigate entanglement spreading and excitation transport within quantum complex networks. For this purpose, we adopt a description in which network nodes represent sites of the physical system (e.g. qubits), and links represent pairs of sites coupled by the Hamiltonian [Bentsen et al., Phys Rev Lett (2019)]. We consider a random tunable-range model [Gori et al., Phys Rev E (2017)] which, in terms of only two parameters, encompasses a large variety of graphs and reduces, in limiting cases, to well-studied benchmarks. Then, it is equipped with a paradigmatic Hamiltonian consisting of hopping terms on the network links. Building on known results [Gori et al., Phys Rev E (2017)], we first study the classical case, providing new semi-analytical estimates. We then address quantum dynamics: For the single-particle case, we analyse the asymptotic entanglement entropy between two halves of the system, computed by Exact Diagonalization. Surprisingly, we observe a new localization region for intermediate ranges of the network links. To understand this phenomenon, we introduce an approximate one-dimensional model and try to relate dynamics to network measures. The analysis is extended to half-filling using fermionic Gaussian states, obtaining results close to the single-particle ones that raise new interesting questions for future work. It is concluded that this model exhibits non-trivial behaviour, especially in the region where unexpected localization occurs. We infer that this is determined by the interplay between network features and quantum effects, possibly induced by structural disorder.
We introduce an elementary quantum system consisting of a set of spins on a graph and a particle hopping between its nodes. The quantum state is build sequentially, applying a unitary transformation that couples neighboring spins and, at each node, couples the local spin with the particle. We observe the relaxation of the system toward a stationary paramagnetic or ferromagnetic state and demonstrate that it is related to eigenvectors thermalization and random matrix statistics. The relation between these macroscopic properties and interaction generated entanglement is discussed.
We introduce a model of a quantum walk on a graph in which a particle jumps between neighboring nodes and interacts with independent spins sitting on the edges. Entanglement propagates with the walker. We apply this model to the case of a one-dimensional lattice to investigate its magnetic and entanglement properties. In the continuum limit, we recover a Landau-Lifshitz equation that describes the precession of spins. A rich dynamics is observed, with regimes of particle propagation and localization, together with spin oscillations and relaxation. Entanglement of the asymptotic states follows a volume law for most parameters (the coin rotation angle and the particle-spin coupling).
The valence-bond structure of spin-1/2 Heisenberg antiferromagnets is closely related to quantum entanglement. We investigate measures of entanglement entropy based on transition graphs, which characterize state overlaps in the overcomplete valence-bond basis. The transition graphs can be generated using projector Monte Carlo simulations of ground states of specific Hamiltonians or using importance sampling of valence-bond configurations of amplitude-product states. We consider definitions of entanglement entropy based on the bonds or loops shared by two subsystems (bipartite entanglement). Results for the bond-based definition agree with a previously studied definition using valence-bond wave functions (instead of the transition graphs, which involve two states). For the one-dimensional Heisenberg chain, with uniform or random coupling constants, the prefactor of the logarithmic divergence with the size of the smaller subsystem agrees with exact results. For the ground state of the two-dimensional Heisenberg model (and also N\'eel-ordered amplitude-product states), there is a similar multiplicative violation of the area law. In contrast, the loop-based entropy obeys the area law in two dimensions while still violating it in one dimension---both behaviors in accord with expectations for proper measures of entanglement entropy.
Entanglement entropy is a powerful tool for characterizing spin liquids and other systems with long-range entanglement. For gapped spin liquids, this problem has been well-studied, finding a universal subleading constant term in the entanglement entropy, known as the topological entanglement entropy. In this paper, the idea of topological entanglement entropy is generalized to the gapless U(1) spin liquid. First, it is shown that the entanglement entropy separates into two pieces, one due to emergent charges of the theory (assumed to be gapped) and one due to the emergent gapless gauge mode. Both give universal subheading contributions to the entanglement entropy, which are logarithmic in the size of the region considered. The photon contribution is due to the local physics of the gapless mode. The particle contribution, on the other hand, is a topological term, generalizing the notion of topological entanglement entropy. A construction is presented which can isolate this topological piece by eliminating the photon contribution and other local terms, such as the area law. This topological term in the entanglement entropy can serve as a powerful diagnostic tool for U(1) spin liquids.
We introduce a mapping between graphs and pure quantum bipartite states and show that the associated entanglement entropy conveys non-trivial information about the structure of the graph. Our primary goal is to investigate the family of random graphs known as complex networks. In the case of classical random graphs, we derive an analytic expression for the averaged entanglement entropy while for general complex networks we rely on numerics. For a large number of nodes n we find a scaling where both the prefactor c and the sub-leading O (1) term g e are characteristic of the different classes of complex networks. In particular, g e encodes topological features of the graphs and is named network topological entropy. Our results suggest that quantum entanglement may provide a powerful tool for the analysis of large complex networks with non-trivial topological properties.