Fracton Topological Phases

• Fracton topological phases are three-dimensional gapped quantum states characterized by immobile fractons and subdimensional excitations with constrained mobility.
• They arise from diverse constructions like coupled-layer, lattice gauge theory, parton, and defect networks, leading to subextensive ground-state degeneracies.
• Field-theoretic models using higher-rank tensor gauge fields and multipole conservation laws explain their exotic transport, braiding phenomena, and quantum phase transitions.

Fracton topological phases are three-dimensional gapped quantum states of matter whose fundamental excitations are subject to extreme subdimensional mobility constraints: while some (fractons) are strictly immobile as isolated entities, others (lineons, planons) can only propagate along one-dimensional lines or within planes. These phases exhibit unconventional features far beyond the landscape of conventional topological order, including ground-state degeneracies that scale subextensively with system size, fusion and braiding phenomena not captured by standard modular tensor category frameworks, and responses linked to higher-rank or multipole conservation laws. They arise through a variety of constructions—coupled-layer, parton-based, generalized lattice gauge theory, and defect-network approaches—and are intimately related to subsystem symmetries and emergent higher-form conservation laws.

1. Defining Properties and Classification

Fracton topological phases are characterized by:

• Subdimensional Excitations:
  • Fractons: point-like quasiparticles that are immobile in isolation; any attempt to move an individual fracton necessitates the creation of additional excitations.
  • Lineons: excitations constrained to move only along specific straight lines (often associated with intersection axes of layered subsystems).
  • Planons: excitations free to move within planes, but not out of them.

These mobility restrictions derive from the presence of subsystem symmetries—symmetries acting on rigid submanifolds such as lines or planes—whose conservation laws enforce the confinement of certain charges within those manifolds (Devakul et al., 2017, Hsieh et al., 2017, Vijay, 2017, Slagle et al., 2017, Ebisu et al., 2024, You et al., 2019, Vijay et al., 2016).

Fracton models naturally separate into two broad classes:

• Type-I Fracton Order: Excitations are generated at the corners of string or membrane operators; logical operators have string or membrane support; GSD grows linearly (or as $L$ on an $L^3$ torus). Examples: X-cube model, checkerboard model, Majorana cube model (Devakul et al., 2017, Hsieh et al., 2017, Vijay, 2017).
• Type-II Fracton Order: Excitations are generated only by operators with support on fractal subsets of the lattice. Logical operators cannot be deformed into strings or membranes. The ground-state degeneracy exhibits large spikes for special system sizes (e.g., $L=2^n$ for Haah’s code) (Hsieh et al., 2017, Ma et al., 2017, Aasen et al., 2020).

The hallmark of both classes is a ground-state degeneracy (GSD) that reflects the underlying subsystem symmetry—a linear or subextensive (e.g., $2^{O(L)}$ rather than $2^{O(L^3)}$) scaling with system size.

2. Model Constructions and Hamiltonians

Coupled-Layer and Lattice Gauge Theory Approaches

Layer constructions stack 2D topological phases (e.g., three families of toric codes in the $xy$, $yz$, $xz$ planes) and introduce interlayer coupling terms that drive the condensation of extended objects such as “$p$-strings” (built from $m$-particles or loop-fluxes) (Vijay, 2017, Ma et al., 2017, Gorantla et al., 23 Sep 2025, Zhu et al., 2022). In the strong-coupling limit, this procedure leads to the X-cube Hamiltonian:

$H_{\rm X{\mbox{-}}cube} = -\sum_v\sum_{i\in\{x,y,z\}}A_{v,i} - \sum_c B_c$

where $$A_{v,i} = \prod_{\ell \in \mathrm{star}_i(v)} X_\ell$$ (product of $X$'s in plane $i$), $B_c = \prod_{\ell \in \partial c} Z_\ell$ (cube term, product of $Z$'s on cube edges). The solvable ground state is an equal amplitude superposition of closed membrane or loop configurations (Vijay, 2017, Ma et al., 2017, Zhu et al., 2022).

Generalized Lattice Gauge Theory

Subsystem symmetries are gauged by introducing new degrees of freedom (nexus fields) associated with local interaction terms defined over planes or fractal subsets (Vijay et al., 2016, Ebisu et al., 2024). The resulting models possess generalized Gauss law constraints at each site and “flux” type operators, leading to nontrivial commutation relations and new conservation laws.

Parton and Majorana Constructions

Partonic representations decompose physical degrees of freedom (Majorana fermions or spins) into clusters subject to overlapping gauge constraints—planar for type-I, fractal for type-II. Non-interacting parton models can reproduce the quantum numbers and logical operator structure of known fracton codes; commuting-projector interacting parton models yield exactly solvable type-I and type-II fracton phases (Hsieh et al., 2017, You et al., 2018, Halász et al., 2017).

Defect Network Framework

Topological defect networks stratify space into domains (strata) carrying bulk TQFTs connected by defects that condense specific sets of anyons (Lagrangian algebras). The fusion and condensation rules at defect intersections enforce the mobility constraints and generate the subdimensional excitation structure—immobile charges, subdimensional strings, and fractal excitation patterns (Aasen et al., 2020).

3. Excitations, Mobility Constraints, and Braiding

Fracton phases defy the standard anyon paradigm for quasiparticle mobility and statistics.

Table: Excitation Classification in Representative Models

Excitation Type	Creation Operator Support	Allowed Mobility	Example Model
Fracton	Membrane corners / fractal support	None (isolated)	X-cube, Haah's code
Lineon	String endpoints	Along a line	X-cube
Planon	Membrane edge	In a 2D plane	Checkerboard, X-cube bound pairs
Non-Abelian Fracton	Fractal/cage operators with twist	None (immobile)	Gauged Haah's code/X-cube (Bulmash et al., 2019)

• Braiding in fracton phases is subtle and generally geometry-dependent and nonreciprocal in the bulk. Statistical phases depend on the spatial arrangement and type of excitations, and bulk cage operators may not be uniquely defined (Bulmash et al., 2018).
• At boundaries, braiding becomes well-defined through reciprocal boundary half-cage operators (BHCs), leading to generalized “boundary Lagrangian subgroup” classification for gapped boundaries (Bulmash et al., 2018).
• Non-Abelian immobile fractons and non-Abelian string excitations can be realized by gauging layer-exchange symmetries or in certain defect networks, leading to robust degeneracies and fusion structures not present in purely Abelian phases (Bulmash et al., 2019, Aasen et al., 2020).

4. Entanglement and Topological Invariants

Standard 2D topological orders are classified by a constant topological entanglement entropy (TEE) extracted from the subleading term in region entanglement entropy scaling. In contrast, fracton phases are characterized by a universal linear (in subsystem size) correction to the area law:

$$S(A) = \alpha|\partial A| - \gamma_{\mathrm{lin}}L_A - \gamma_0 + \dots$$

where $L_A$ is the linear size of region $A$, and $\gamma_{\mathrm{lin}}$ defines the fracton topological entanglement entropy.

• For the X-cube model, $\gamma_{\mathrm{lin}}=1$ (ABC prescription) or $2$ (PQWT construction), reflecting the number of nonlocal stabilizers associated with each foliation (Ma et al., 2017).
• Haah’s code (type-II) shows a still larger $\gamma_{\mathrm{lin}}$ in certain prescriptions, consistent with a “fractal condensate” wavefunction structure.
• The linear correction is robust to arbitrary local perturbations, as proven via Schrieffer–Wolff local uni­tary transformations (Ma et al., 2017).

These properties indicate that fracton phases lack a conventional continuum TQFT description but are distinguished by multiplet topological invariants $(\gamma_x,\gamma_y,\gamma_z)$ tied to the spatial foliation structure or fractal geometry.

5. Field-Theoretic Descriptions and Multipole Laws

Fracton phases are unified by effective low-energy field theories built from higher-rank tensor gauge fields and multipole conservation laws:

• Multipole Chern-Simons and BF Theories:

In “Multipolar Topological Field Theories,” rank-2 gauge fields describe phases with conserved dipole or quadrupole moments and subsystem symmetries. The effective action may contain, for quadrupole (2D) cases:

$$S_{\rm quad} = \frac{\theta}{2\pi} \int \left(\partial_x\partial_yA_0 - \partial_tA_{xy}\right)$$

And in 3D, dipolar Chern-Simons terms:

$$S_{\rm dip-CS} = \frac{k}{4\pi} \int \left[ A_z E_{xy} + A_{xy} E_z - A_0 B \right]$$

These predict: - Immobility of isolated fractons, - Subdimensional transport of multipoles, - Anomalous edge/corner charges and chiral hinge currents (as found in higher-order topological insulators) (You et al., 2019, Ebisu et al., 2024).

• Subsystem Symmetries and Higher-Form Gauging:

The relationship between global, subsystem, and higher-form symmetries is elucidated in constructions where X-cube order (and generalizations) arises from gauging the 1-form symmetry defects in a foliated stack of 2D topological orders. The gauging web relates fracton order, enriched toric code, and SPT phases classified by subsystem and higher-form symmetry extensions (Gorantla et al., 23 Sep 2025).

6. Quantum Phase Transitions and Criticality

Fracton topological orders exhibit unconventional phase transitions, including:

• Fracton Confinement Transition:

Tuning parameters (e.g., transverse field, deformation strength in tensor-network representations) through a critical point can destroy the subextensive ground-state degeneracy, confining fractons. For type-I fracton orders, the critical point for a field creating lineons is universal: $v_c = w/4$, where $w$ is the pair creation energy (Poon et al., 2020, Zhu et al., 2022). The transition can be first-order (for finite $N$ in $\mathbb{Z}_N$ generalizations) or continuous in the $N\to\infty$ limit, in which case the system exhibits a line of conformal quantum critical points with deconfined fracton states at criticality.

• Partial Deconfinement and Diagnostic Operators:

In fracton phases, standard Wilson-loop diagnostics fail due to subdimensional mobility. Instead, generalized “Wil­son ribbons” and horseshoe operators (products over ribbons or open membranes with matter insertions at ends) detect partial deconfinement—free mobility of bound pairs within their allowed subspaces. The ratio $\mathcal{R}(L)$ of open/closed ribbon expectation values sharply distinguishes between fracton (vanishing with length) and confined (nonzero) phases (Devakul et al., 2017).

• Role of Layer Construction and Dualities:

Many fracton phases are proximate to stacks of lower-dimensional topological orders, and transitions can often be mapped via duality to critical phenomena in classical gauge theories, transverse-field Ising models, or plaquette clock models, facilitating theoretical and numerical analysis (Vijay, 2017, Slagle et al., 2017, Zhu et al., 2022).

7. Generalizations, Non-Abelian Fractons, and Defect Networks

Beyond Abelian Phases

• Gauging layer-exchange or internal symmetries in multiple-copy fracton models generates non-Abelian immobile fractons whose internal degeneracies and fusion rules depend on global geometry—quantum dimensions become geometry-dependent and subdimensional (Bulmash et al., 2019, Aasen et al., 2020).
• Defect network formalism provides a general organization principle: all gapped fracton phases (including type-II) are realized as stratified TQFTs with appropriate defect data, with mobility restrictions enforced by the condensation and braiding of anyons trapped at lower-dimensional strata. This formalism unifies Abelian, non-Abelian, and fractal-type fracton phases and constrains, for example, the nonexistence of stable type-II fracton order in purely $2+1$D gapped systems (Aasen et al., 2020).

Boundary Theory and Classification

• A boundary Lagrangian subgroup (BLS), defined via boundary half-cage operators and reciprocal boundary braiding, appears necessary and sufficient to characterize gapped boundaries of Abelian type-I fracton models—mirroring the modular tensor category formalism in 2D (Bulmash et al., 2018).
• Higher-order gapped boundaries (hinges, corners) and their classification are naturally included in the defect network framework, building on the classification of gapped boundaries in 3+1D TQFTs (Aasen et al., 2020).

---

Fracton topological phases constitute a broad and structurally rich class of three-dimensional quantum matter, unified by the interplay of higher-form and subsystem symmetries, multipole conservation laws, and generalized defect or parton constructions. They encode new forms of localization, braiding, and quantum information storage, and provide a laboratory for exploring non-TQFT order, exotic quantum criticality, and subsystem-protected quantum phenomena. The algebraic and field-theoretic frameworks developed for fracton phases increasingly point toward a new paradigm for classifying and realizing topological order beyond conventional anyonic systems.