arXiv:2609.19282
Non-Abelian Anyon Condensation: a Path-Integral Monte Carlo Approach
Rafael Flores-Calderón, Frank Pollmann, Michael Knap
Abstract
Transitions out of non-Abelian topological order are difficult to describe in microscopic quantum models with numerical methods that remain tractable at large scales. We develop a sign-free path-integral framework for Kitaev quantum doubles $\mathcal D(G)$ by organizing single-link perturbations in terms of non-invertible electric and magnetic 1-form symmetries that proliferate distinct anyon species. For any finite group $G$, an exact $\textit{matterization}$ isometry introduces vertex degrees of freedom and maps the link-only model onto a $G$ gauge--Higgs theory. Furthermore, the 1-form symmetries of the fixed point allow us to define generalized Fredenhagen--Marcu order parameters that become finite when the corresponding anyons condense. For $G = S_3$, quantum Monte Carlo simulations show that proliferating a non-Abelian electric anyon drives a first-order transition in which all nontrivial electric anyons condense. In the purely magnetic limit, the model reduces to a $(2+1)$D pure $G$ gauge theory; for $G=S_3$, it exhibits a first-order confinement transition, diagnosed by the onset of a Wilson-loop area law and the restoration of an emergent magnetic 1-form symmetry. These results provide a unified numerical framework for non-Abelian anyon condensation, confinement, and generalized symmetry breaking.
AI Summary
The paper studies how local perturbations drive condensation and confinement out of non-Abelian quantum-double topological order. It constructs an exact “matterization” isometry for any finite group $G$, introducing vertex variables and mapping perturbed $\mathcal D(G)$ link models onto gauge–Higgs theories with a physical-subspace Gauss-law constraint. A Trotterized path integral gives a $(2+1)$-dimensional classical model with nonnegative weights for real, nonnegative magnetic class-sum amplitudes and Hermitian perturbations. Electric perturbations enter as representation-resolved Higgs couplings. The construction also supplies irrep-resolved Fredenhagen–Marcu (FM) string observables to diagnose electric condensation. For $G=S_3$, heat-bath Monte Carlo simulations examine two routes out of topological order. With equal magnetic amplitudes $\lambda_D=\lambda_F=\lambda$ and no electric perturbation, confinement occurs near $\lambda/\kappa\simeq0.075$: Wilson loops acquire an area-law contribution and noncontractible loops lose flux-sector dependence. Energy discontinuities and hot–cold-start coexistence support a first-order transition. At $J/\kappa=10$, $\lambda/\kappa=0.01$, and $\lambda_B=0$, increasing the non-Abelian electric field produces simultaneous onsets of the $E$- and $A_2$-sector FM responses near $\lambda_C/\kappa\simeq1.75$. Size-dependent sharpening and metastable branches support direct first-order condensation into a trivial Higgs regime. The numerical conclusions rely on finite sizes and imaginary-time discretization; quoted branch errors exclude tunnelling, finite-size drift, and Trotter uncertainties. FM ratios become statistically unresolved where closed-loop denominators are too small. Independently tuned fields and possible intermediate $\mathcal D(\mathbb Z_2)$ phases remain unexplored.
Error found: The End Matter incorrectly gives an O((λ/κ)^4) remainder after the quadratic deconfined Wilson-loop perimeter coefficient: on the uniform S3 magnetic-hopping line, same-link two-hop processes generate a nonzero cubic contribution, giving αΓ=(3/2)(λ/κ)^2+6(λ/κ)^3+O((λ/κ)^4) for either nontrivial irrep.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.