Gapped Fracton Phases: Subdimensional Quantum Orders

• Gapped fracton phases are 3D gapped quantum states characterized by subdimensional excitations, robust ground-state degeneracy, and non-liquid entanglement structures.
• They are realized in exactly solvable models like the X-cube model and Haah’s code, which use commuting projector Hamiltonians and fractal operator techniques.
• The unique fusion rules and geometry-dependent degeneracy in these phases offer new paradigms for quantum information storage and phase transition studies.

Gapped fracton phases are three-dimensional gapped quantum phases distinguished by emergent topological order with subdimensional excitation mobility and a robust energy gap. These phases are characterized by the existence of point-like or higher-form excitations (fractons, lineons, planons) whose motion is strongly restricted; a ground-state degeneracy that grows with system size in a nontrivial, typically subextensive manner; and entanglement signatures that cannot be accounted for by conventional topological quantum field theory (TQFT). Gapped fracton phases cannot be smoothly connected to trivial or liquid topological phases without either closing the gap or breaking the underlying lattice or subsystem symmetries. They constitute a fundamentally new class of non-liquid topological states, with intricate fusion and braiding properties, robust entanglement signatures, and a rich web of models and classification frameworks.

1. Defining Properties and Microscopic Realizations

Gapped fracton phases are distinguished by the following key features (Shirley et al., 2018, Wen, 2020, Bulmash et al., 2019):

• Restricted Mobility: Excitations are classified by the dimensionality of their allowed motion:
  • Fractons: Strictly immobile, cannot move without creating additional excitations.
  • Lineons: Mobile only along one-dimensional lines.
  • Planons: Mobile within two-dimensional planes.
• Ground-State Degeneracy (GSD): The GSD typically grows sub-extensively with the linear system size and depends on both the topology and the lattice geometry. For example, in the X-cube model, GSD$\sim2^{3L-3}$ for $L\times L\times L$ cubic systems.
• Entanglement Structure: These phases exhibit a linear-in-length correction to the area law in entanglement entropy, distinct from the universal constant correction in 2D topological order (Ma et al., 2017).
• Commuting Projector Hamiltonians: Many paradigmatic models are exactly solvable stabilizer codes, such as the X-cube and Haah’s code (Ma et al., 2017).
• No TQFT Description: Fracton phases are not described by conventional continuum TQFT, as their entanglement and excitation content retains information about the underlying "cellular" or foliation structure at all RG scales (Wen, 2020, Shirley et al., 2018).

Table: Hallmarks of Gapped Fracton Phases

Feature	Gapped Fracton Phase	Conventional Topological Phase
Excitation Mobility	Subdimensional (fractons etc.)	Fully mobile anyons/loops
GSD Scaling	Grows with $L$ (subextensive)	Constant with $L$
Entanglement Topological Term	Scales linearly in length	Constant (2D $\gamma$)
TQFT Description	Absent	Present
Structure	Cellular/Foliated	Homogeneous liquid

2. Classification and Exactly Solvable Models

Two archetypal gapped fracton phases are (Ma et al., 2017, Shirley et al., 2018, Wen, 2020):

• X-cube Model: Qubits on cubic edges; Hamiltonian composed of vertex and cube stabilizers. Fractons created at cube term violations, lineons at vertex term violations. GSD grows as $2^{3L-3}$.
• Haah’s Code: Two qubits per lattice site, with cubic stabilizers enforcing eight-qubit constraints. Fracton excitations are strictly immobile and created only by fractal operators; GSD has a complicated, fractal function dependence on $L$.

Other models include checkerboard codes, twisted X-cube models, and so-called "type-II" phases (e.g., Haah’s code) where excitations lack even planon or lineon motion (Wen, 2020, Song et al., 2018, Shirley et al., 2019).

Cellular and defect-network constructions systematically generate models with prescribed mobility and ground-state structure by gluing layered 2D topological orders using gapped boundary conditions (Lagrangian algebras) along a cellular decomposition (Wen, 2020, Aasen et al., 2020).

Defect TQFT frameworks generalize these lattice constructions: by condensing sets of bulk anyons or fluxes on lower-dimensional strata (planes, lines, points) embedded in a 3+1D TQFT, one engineers the subdimensional mobility of excitations. All known (type-I and type-II) gapped fracton models, including more exotic non-Abelian examples, can be obtained as defect networks in 3+1D lattice gauge theories (Aasen et al., 2020).

3. Fusion, Braiding, and Superselection Sectors

Fusion and statistics in gapped fracton phases are controlled by translation-enriched fusion theory (Pai et al., 2019, Song et al., 2018). The superselection sectors of excitations can be encoded as modules over the translation group ring $R=\mathbb{Z}[\mathbb{Z}^3]$, explicitly capturing their mobility restrictions.

• Fusion Structure: The fusion group is typically Abelian, but in twisted or hybrid/nonsolvable models, inextricably non-Abelian fracton excitations can emerge, featuring quantum dimension $d>1$ not reducible by fusing with higher-mobility excitations (Song et al., 2018, Bulmash et al., 2019, Tantivasadakarn et al., 2021).
• Braiding: Subdimensional mobility restricts the existence of conventional braiding, but nontrivial statistical phases arise when certain membrane or cage operators are moved around each other in higher-dimensional processes. In twisted models, "inextricably non-Abelian" statistical processes persist even when all mobile degrees of freedom are added or removed (Song et al., 2018).
• Geometry-Dependent Degeneracy: The superselection and fusion structure of multiple fractons depends not only on their number but also on their relative positions—the degeneracy is geometry-sensitive and encoded in the membrane/cage operator algebra (Bulmash et al., 2019).

4. Field-Theoretic and Tensor-Network Descriptions

Gapped fracton phases admit highly nontrivial field-theoretic and tensor-network formulations reflecting their non-liquid, UV-sensitive entanglement (Spieler, 2023, Shirley et al., 2018, Zhu et al., 2022):

• Foliated Field Theory: Employs a continuum Lagrangian with gauge fields defined only along a set of fixed, mutually intersecting 2D foliation planes. The action includes BF-like terms coupling standard and leaf-restricted gauge fields, correctly reproducing the excitation and fusion content (Spieler, 2023).
• Exotic Tensor Gauge Theory: Constraints from the foliated formalism yield higher-rank symmetric tensor gauge fields, with generalized (often unconventional) gauge transformations. The fusion and holonomy structure of fracton operators are encoded in symmetric tensor fields, whose Gauss laws enforce the restricted mobility (Spieler, 2023, Ma et al., 2018).
• Tensor-Network States: Exact PEPS or projected entangled pair state representations for both the X-cube and $\mathbb{Z}_N$ generalizations permit direct calculation of ground-state wavefunctions, entanglement entropy, and realization of confinement transitions. Order parameters such as membrane condensates and fracton/loop confinement lengths are naturally accessible (Zhu et al., 2022).

5. Entanglement, Phase Structure, and Transitions

• Entanglement Diagnostics: The area law for entanglement entropy is supplemented by a robust, linear-in-length topological correction—universal within a given phase and invariant under all local, gapped perturbations (Ma et al., 2017). Multipartite wireframe invariants cleanly distinguish trivial phases, 3D liquid topological orders, and nontrivial gapped fracton phases (Shirley et al., 2018, Shirley et al., 2018).

Model	$\gamma_{\rm frac}^{\rm ABC}$	$\gamma_{\rm frac}^{\rm PQWT}$
X-cube	$+1$ (per cube)	$+2$ (per plane)
Haah's code	$+2$ (per cube)	$+4$ (per orientation)

• Phase Transitions: Gapped fracton phases can undergo anyon condensation transitions to conventional 3D topological orders or vice versa via condensation of, e.g., planon or lineon composites (Tantivasadakarn et al., 2021). For $\mathbb{Z}_N$ X-cube models, a line of weakly first-order fracton confinement transitions terminates at a continuous fracton quantum critical point (DQCP) as $N\to\infty$ (Zhu et al., 2022). Hybrid fracton orders interpolate between topological and fractonic regimes, with phase transitions described by anyon condensation or Higgs-type mechanisms (Tantivasadakarn et al., 2021, Tantivasadakarn et al., 2021, Spieler, 2023).

6. Classification Schemes and Extensions

Several complementary classification approaches for gapped fracton phases have been proposed:

• Defect Homology and Symmetry Defect Condensation: The phase space of gapped fracton orders can be systematically constructed via integer-linear constraint equations for the phase factors decorating local projectors, generalizing the string-net equations for 2D topological order (Tantivasadakarn et al., 2019).
• Cellular Topological States: Building up 3D fracton phases from networks of coupled 2D topological orders and gapped 1D boundaries (cellular TQFTs) encodes the fixed-point wavefunctions and mobility/fusion constraints (Wen, 2020).
• Hybrid Orders and Non-Abelian Extensions: Fracton models with both topologically mobile and fractonic sectors emerge from gauging both global and subsystem symmetries of a finite group $G$ with an Abelian normal subgroup $N$. The resulting universal data—excitation content, mobility, fusion, and braiding—are fully determined by $G, N$, and the foliation geometry (Tantivasadakarn et al., 2021, Tantivasadakarn et al., 2021).
• Defect TQFT and Membrane-Net Construction: All known gapped fracton phases arise as topological defect networks in stratified 3+1D TQFTs, via condensation of extended defect layers to enforce mobility constraints (Aasen et al., 2020). This framework unifies Abelian, non-Abelian, hybrid, and twisted gapped fracton orders.

7. Interfaces, Boundaries, and Physical Realizations

Boundary and interface phenomena in gapped fracton phases reveal new defect and condensation behaviors (Hsin et al., 2023):

• Gapped Boundaries: Gapped boundaries can be constructed by condensing maximal sets of mutually bosonic excitations consistent with the subsystem and foliation constraints. There exist both undecorated and "decorated" boundaries (with additional Chern–Simons-like boundary actions) in, e.g., the X-cube model (Hsin et al., 2023).
• Interfaces with Topological Orders: Physical interfaces between gapped fracton phases and conventional 3D topological orders (e.g., the toric code) can be systematically classified and constructed, with duality defects generalizing Kramers–Wannier and electromagnetic dualities to the foliated 3+1D setting.
• Experimental and Theoretical Relevance: These phases provide new paradigms for robust quantum information storage, for understanding the limits of TQFT, and for the classification of unmappable entangled matter in three or more spatial dimensions.

---

Gapped fracton phases represent a profound generalization of topological order, characterized by non-liquid, UV-sensitive topological phenomena, subdimensional constrained dynamics, and robust ground-state and entanglement structures that distinguish them from all liquid TQFT phases (Shirley et al., 2018, Aasen et al., 2020, Wen, 2020, Tantivasadakarn et al., 2021, Bulmash et al., 2019).