1: \begin{abstract}

2: We construct a family of exactly solvable spin models that illustrate a novel mechanism for fractionalization in topologically ordered phases, dubbed the string flux mechanism.  The essential idea is that an anyon of a topological phase can be endowed with fractional quantum numbers when the string attached to it slides over a background pattern of flux in the ground state.  The string flux models that illustrate this mechanism are $\z_n$ quantum double models defined on specially constructed $d$-dimensional lattices, and possess $\z_n$ topological order for $d \geq 2$.  The models have a unitary, internal symmetry $G$, where $G$ is an arbitrary finite group. The simplest string flux model is a $\zz$ toric code defined on a bilayer square lattice, where $G = \zz$ is layer-exchange symmetry.  In general, by varying the pattern of $\z_n$ flux in the ground state, any desired fractionalization class [element of $H^2(G, \z_n)$] can be realized for the $\z_n$ charge excitations.  While the string flux models are not gauge theories, they map to $\z_n$ gauge theories in a certain limit, where they follow a novel \emph{magnetic} route for the emergence of low-energy gauge structure.  The models are analyzed by studying the action of $G$ symmetry on $\z_n$ charge excitations, and by gauging the $G$ symmetry.  The latter analysis confirms that distinct fractionalization classes give rise to distinct quantum phases, except that classes $[\omega], [\omega]^{-1} \in H^2(G, \z_n)$ give rise to the same phase.  We conclude with a discussion of open issues and future directions.

3: \end{abstract}

4: