Fracton Models in Quantum Matter

• Fracton models are quantum spin systems characterized by point excitations with severely limited mobility due to strict conservation laws.
• They exhibit subextensive ground-state degeneracy and novel entanglement structures that challenge conventional 2D topological order.
• Diverse lattice and field-theoretic approaches, including coupled-layer constructions and tensor gauge theories, reveal new paradigms for quantum memory and non-Abelian phases.

Fracton models are a class of gapped quantum spin systems in three or more spatial dimensions supporting point-like excitations—fractons—whose mobility is fundamentally and often severely restricted by microscopic or emergent conservation laws. These models have catalyzed major advances in the understanding of quantum phases and topological order far beyond the two-dimensional landscape, revealing novel paradigms for topological protection, information storage, entanglement structure, and quantum criticality. Fracton phases are generally characterized by subextensive ground-state degeneracy, subsystem symmetry or foliation structure, and unconventional dynamics including “glassy” slow relaxation and non-ergodicity.

1. Fundamental Properties and Classification

Defining signatures. Fracton order is diagnosed by two core properties:

• The ground-state degeneracy (GSD) grows subextensively with the linear size $L$ of the system, often as $\log_2 \mathrm{GSD} \sim L^\alpha$ for $0<\alpha\le D$ in $D$ dimensions.
• Elementary excitations—fractons—are point-like objects whose individual mobility is strictly constrained. Depending on the model, one also finds excitations with dimension-1 or dimension-2 restricted mobility, “lineons” or “planons.”

Types of fracton order.

• Type I: Point excitations (fractons) that can move only when bound into multiplets—for instance, lineons that move along lines or planons in planes. The X-cube model is the canonical type-I example (Ma et al., 2017, Pretko et al., 2020).
• Type II: Point excitations are strictly immobile in isolation, with no lower-dimensional mobile composites. Haah's cubic code exemplifies this class (Ma et al., 2017, Pretko et al., 2020).

Product code constructions. The hypergraph-product (HGP) code yields CSS stabilizer Hamiltonians whose parameters and excitation phenomenology are governed by the properties of two input classical codes $H_1, H_2$. Specifically, fracton order emerges when the seeds are (i) rank-deficient, (ii) confining, and (iii) isolable, ensuring an extensive kernel or cokernel, protection against low-weight errors, and the ability to separate syndromes—capturing the nonlocality and mobility restrictions characteristic of fracton phases (Tan et al., 2023).

2. Lattice Models and Coupled-Layer Constructions

X-cube and related models.

• The X-cube model is defined on a cubic lattice with a qubit per edge and mutual commuting stabilizers for each cube and each coordinate plane at every vertex. It realizes fractons as cube excitations and lineons localized to axes, with GSD scaling as $2^{6L-3}$ on the $L^3$ torus (Ma et al., 2017, Shirley et al., 2017, Pai et al., 2019, Pretko et al., 2020).
• Checkerboard models, semionic X-cube models, and Four Color Cube models are obtained via coupled-layer arrangements or p-membrane condensation, generating richer excitation content and more complex GSD scaling (Ma et al., 2017, Wickenden et al., 2024).

Parton approaches. Two classes of parton constructions—non-interacting (Majorana) and interacting (bosonic)—allow the systematic realization of both type-I and type-II fracton codes, with the gauge group structure directly reflected in the ground-state degeneracy and fractal logical operators (Hsieh et al., 2017).

Anisotropic and cage-net models.

• Coupled-layer methods can create highly anisotropic fracton models by stacking and condensing anyon pairs across 2D topological layers. Resulting models manifest subdimensional excitations whose mobility is tied to the underlying layer geometry (Fuji, 2019).
• Cage-net models constructed from stacking and condensing flux-strings in 2D string-net layers yield three-dimensionally intrinsic non-Abelian subdimensional anyons—objects that cannot be realized as bound states of lower-dimensional anyons (Prem et al., 2018).

Twisted fracton models. Twisted generalizations parametrize fracton phases using slice-wise Dijkgraaf–Witten-type cohomology, allowing inextricably non-Abelian fractons and ground-state degeneracy with intricate dependence on system size and boundary conditions (Song et al., 2018).

3. Field-Theoretic Descriptions and Dualities

Higher-moment conservation and gauge structure.

• Fracton phases emerge as ground states of symmetric tensor gauge theories, such as the rank-2 $U(1)$ scalar-charge theory, where conservation of charge and dipole moment enforces fractonic immobility (Pretko et al., 2020, Gromov et al., 2022, Pretko et al., 2017).
• The “fracton-elasticity duality” establishes a precise correspondence between the elasticity theory of 2D (and higher) quantum crystals and fracton tensor gauge theory—disclinations correspond to fractons, dislocations to dipoles with restricted glide motion, and phonons to gapless gauge photons (Pretko et al., 2017, Gromov et al., 2022).

Continuum and foliation structure. Fracton phases challenge standard TQFT notions. Ground-state degeneracy and operator structure depend not only on the global topology but on a foliation or layering—explicit in lattice model cellulations and captured by the classification of foliated fracton phases (Shirley et al., 2017, Pretko et al., 2020, Wickenden et al., 2024).

Gauged multipole symmetries. Beyond conventional gauge theory, fracton models can be viewed as the result of gauging extended multipole symmetries. By selecting polynomials of spatial coordinates to conserve, an infinite landscape of rank-$k$ tensor gauge theories with tunable fractonic mobility constraints can be systematically constructed (Gromov et al., 2022).

4. Excitations, Fusion Theory, and Statistics

Superselection structure.

• The set of point excitations forms a module over lattice translations; the stabilizer subgroup identifies mobile, lineon, planon, or fractonic sectors. Planon-modular (p-modular) fusion theory has emerged as an organizational framework for the classification of type-I models, where all nonzero sectors can be detected by braiding with a planon (Wickenden et al., 2024).
• Fusion rules and mobility constraints are tightly correlated—for instance, in the X-cube model only pairs of fractons in a plane can form mobile planons, while lineons are strictly bound to lattice axes (Pai et al., 2019).

Statistical processes and invariants. Braiding and exchange statistics in fracton phases require an analysis of spatially separated local moves with restricted deformations, often displaying braiding statistics distinct from 2D topological phases. Notable distinctions arise between, for example, the X-cube and semionic X-cube models via lineon exchange phases (Pai et al., 2019, Song et al., 2018, Wickenden et al., 2024).

Non-Abelian fracton phases. By introducing categorical twists or constructing models with non-Abelian anyon input, fracton phases supporting non-Abelian lineons or inextricably non-Abelian fractons have been realized. Braiding and fusion properties can remain non-Abelian even after accounting for all mobile excitation fusion, an inherently three-dimensional phenomenon (Prem et al., 2018, Song et al., 2018).

5. Entanglement, Dynamics, and Thermal Properties

Topological entanglement entropy.

• Fracton models present novel entanglement entropy structure: the topological contribution scales linearly with subsystem size, reflecting the global constraints arising from nonlocal stabilizers and foliation (Ma et al., 2017).
• This linear scaling is a robust marker, persisting under local perturbations and offering a probe for fracton order in many-body localized settings.

Quantum dynamics and glassiness.

• The restricted microscopic dynamics in fracton models leads to anomalously slow relaxation, glassy behavior, and subdiffusive or nonergodic regimes even in the absence of quenched disorder (Prem et al., 2017, Pretko et al., 2020).
• Type-I fracton models exhibit exponentially suppressed mobility and logarithmic relaxation under thermalization, while type-II models display super-Arrhenius equilibration times and subdiffusion.

Finite-temperature transitions and quantum memory.

• In 3D, fracton topological order is generically destroyed at any nonzero temperature: no-go theorems tie this to subsystem symmetry breaking and many-body localization arguments. In higher dimensions, e.g., the 4D X-cube model, stabilized fracton order at finite temperature and sharp confinement–deconfinement transitions are found (Shen et al., 2021).
• At low temperature, these models can provide classical self-correction, but true quantum self-correcting memory likely requires even higher spatial dimensions or exotic symmetry architectures.

6. Recent Developments and Generalizations

Planon-modular and foliated fracton orders.

• Planon-modular fracton theories have been advanced as a unifying framework for type-I orders, providing algebraic phase invariants based on the ability of planons to detect all sectors via braiding, and supporting a classification under exotic entanglement RG flows, including foliated RG (Wickenden et al., 2024).
• Precisely solvable models including $\mathbb{Z}_n$ X-cube, anisotropic, checkerboard, 4-planar X-cube, and FCC models have been analyzed under this classification, showing detailed relationships among their intrinsic weight invariants, constraint modules, and RG fixed points.

Neural quantum states and computational studies.

• Neural-network-based quantum state representations, such as restricted Boltzmann machines and correlation-enhanced RBMs, have enabled numerically precise studies of fracton models on hundreds of qubits and revealed sharp first-order quantum phase transitions under applied perturbations, inaccessible to conventional Monte Carlo approaches (Machaczek et al., 2024).

Physical realizations and experimental routes.

• The fracton framework has found entry points into experimental and quasi-experimental systems: tilted optical lattices (for dipole conservation), hole-doped antiferromagnets (fracton–polaron correspondence), and arrays of Majorana islands have all been shown to instantiate fractonic phenomenology or provide natural platforms for fracton Hamiltonians (Sous et al., 2019, Gromov et al., 2022, Hsieh et al., 2017).

Holography and gravitational analogies. Fracton models exhibit features reminiscent of gravitational physics (e.g., emergent attraction, holographic entanglement, and black hole analogues in hyperbolic models), suggesting deep connections with quantum information and spacetime geometry (Yan, 2019, Gromov et al., 2022).

7. Outlook and Open Questions

Fracton models have established a new paradigm for topological quantum matter extending far beyond conventional 2D TQFT. Current frontiers include:

• Systematic classification of non-Abelian and type-II order beyond the planon-modular framework.
• Explicit realization in experimental platforms and material settings.
• Formalization of “fracton TQFT” as a higher-category or foliation-based field theory.
• Investigation of their universal dynamical and entanglement properties, and potential quantum information processing capabilities in higher dimensions or under novel symmetry constraints.

The field continues to evolve rapidly, integrating insights from quantum error correction, statistical mechanics, elasticity theory, and category theory into a comprehensive theory of quantum matter dominated by restricted mobility, long-range entanglement, and higher-moment conservation (Pretko et al., 2020, Gromov et al., 2022, Wickenden et al., 2024).