Electric-magnetic duality of lattice systems with topological order
Oliver Buerschaper, Matthias Christandl, Liang Kong, Miguel Aguado
TL;DR
This work identifies electric-magnetic duality as a unifying principle for lattice systems with topological order, extending beyond Abelian toric codes to non-Abelian quantum doubles by placing them within the Hopf-algebra framework $[\mathrm{D}(H)]_\Lambda$ and their duals $[\mathrm{D}(H^*)]_{\Lambda^*}$. It introduces extended string-net models $[\mathrm{ESN}_{(\mathcal{C},\omega)}]_\Lambda$ via a fiber functor $\omega$, connects them to tensor-categorical duals $\mathcal{C}^*$ through Tannaka-Krein duality, and shows EM duality is implemented by invertible domain walls between dual lattice systems. By projecting ESN models to ordinary string-nets, the paper derives a duality between magnetic and electric projections, relates these to Morita equivalences, and identifies the conditions under which all SN models may be accommodated (via weak $C^*$-Hopf algebras). It culminates in a framework where topology is measurable through dual tensor-network states, linking lattice models, tensor categories, and domain-wall physics, with implications for experimental realizations and further dualities among topological phases.
Abstract
We investigate the duality structure of quantum lattice systems with topological order, a collective order also appearing in fractional quantum Hall systems. We define electromagnetic (EM) duality for all of Kitaev's quantum double models based on discrete gauge theories with Abelian and non-Abelian groups, and identify its natural habitat as a new class of topological models based on Hopf algebras. We interpret these as extended string-net models, whereupon Levin and Wen's string-nets, which describe all intrinsic topological orders on the lattice with parity and time-reversal invariance, arise as magnetic and electric projections of the extended models. We conjecture that all string-net models can be extended in an analogous way, using more general algebraic and tensor-categorical structures, such that EM duality continues to hold. We also identify this EM duality with an invertible domain wall. Physical applications include topology measurements in the form of pairs of dual tensor networks.
