Checkerboard Code in Quantum Lattices

• Checkerboard codes are bipartite lattice models defined on square or hexagonal lattices with alternating stabilizers that generalize toric codes.
• They employ split star operators and anisotropic plaquette interactions to realize U(1) and XY extensions, enabling tunable phase transitions.
• Recent studies reveal detailed ground-state degeneracy, fracton excitations, and rigorous bounds on their information-theoretic capacity.

The term “Checkerboard Code” characterizes a class of lattice models and associated quantum error-correcting codes, distinguished by a bipartite (checkerboard) arrangement of local operators or code stabilizers on square or hexagonal lattices. These constructions have found applications in quantum information, the study of fracton and topological phases, and theoretical analysis of code capacity and degeneracy. Recent developments include the U(1) and XY checkerboard toric codes, which generalize the canonical Kitaev toric code by splitting star operators and introducing symmetry-enriched or anisotropic interactions in a checkerboard pattern. The capacity of checkerboard codes on hexagonal lattices and their rigorous bounding methodologies have also attracted interest in coding theory, albeit with limited publicly available results (Deng et al., 2015).

1. Lattice Construction and Operator Geometry

Checkerboard codes are defined on two-dimensional lattices—typically square or hexagonal—where physical degrees of freedom (e.g., spin-½ variables characterized by Pauli matrices $\sigma^x, \sigma^y, \sigma^z$) are organized such that local terms (“stars,” “plaquettes”) form an alternating bipartite structure. In the square-lattice models, stars (four-spin operators) reside on sublattice vertices in a checkerboard fashion, distinguished as $s_1$ and $s_2$ (Vieweg et al., 30 May 2025, Vieweg et al., 2024). The local code stabilizers or Hamiltonian terms alternate between sublattices, enabling anisotropic or symmetry-enriched extensions of the toric code. In the hexagonal-lattice setting, checkerboard codes are constructed analogously, with distinct local operators covering hexagons in an alternating pattern; bounding their information-theoretic capacity remains an active area (Deng et al., 2015).

2. Hamiltonians and Symmetry Structure

The prototypical Hamiltonian for checkerboard toric codes includes:

• Star operators on two sublattices ($s_1, s_2$), each acting on four spins,
• Plaquette (flux) operators $B_p = \prod_{i\in p}\sigma^z_i$ measuring emergent $\mathbb{Z}_2$ flux,
• Interpolating parameters (e.g., $\lambda$, flavor angle $\phi$, anisotropy couplings $J_{s_1}, J_{s_2}$) that tune interactions between star types.

For the U(1) checkerboard toric code (U1TC), stars host U(1)-enriched operators: $$\tilde{A}_s = \sigma_1^+\sigma_2^+\sigma_3^-\sigma_4^- + \sigma_1^+\sigma_2^-\sigma_3^+\sigma_4^- + \sigma_1^+\sigma_2^-\sigma_3^-\sigma_4^+ + \text{h.c.}$$ where $\sigma^\pm = (\sigma^x \pm i\sigma^y)/2$, and the full Hamiltonian is

$$H = -\sum_{s_1} \tilde{A}_{s_1} - \lambda \sum_{s_2} \tilde{A}_{s_2} - \sum_p B_p,$$

endowed with global U(1) (magnetization conservation) and local $\mathbb{Z}_2$ (plaquette) symmetries (Vieweg et al., 30 May 2025).

The XY checkerboard toric code generalizes by mixing x/y flavors: $$H_{\rm XYTC} = -\sum_p B_p - J_{s_1} \sum_{s_1} [ \cos\phi\,A_{s_1}^{(x)} + \sin\phi\,A_{s_1}^{(y)} ] - J_{s_2} \sum_{s_2} [ \cos\phi\,A_{s_2}^{(x)} + \sin\phi\,A_{s_2}^{(y)} ],$$ with independent coupling strengths and flavor angle $\phi$ (Vieweg et al., 2024).

3. Ground-State Degeneracy and Compactification

The degeneracy structure of checkerboard codes is compactification-dependent, arising from global constraints imposed by local stabilizers. For torus compactifications specified by vectors $(a,b)$, degeneracy counts reflect the geometry:

• $0^\circ$ compactification: $\mathrm{GSD}_{0^\circ}(\lambda=0) = 2^{L/2+1} - 1$,
• $45^\circ$ compactification: $\mathrm{GSD}_{45^\circ}(\lambda=0) = 2^{L+1} - 1$,

where L is the linear system size, and degeneracy lies in specific magnetization and Wilson loop sectors. These states correspond to different choices of ground-state superpositions on rows/diagonals of stars, enforced by plaquette constraints (Vieweg et al., 30 May 2025). In the XY checkerboard toric code, analogous sub-extensive degeneracies arise—from $2^L$ in the fracton regime—while topologically ordered phases recover the expected $4^g$ degeneracy for genus g surfaces (Vieweg et al., 2024).

4. Excitations and Fracton Phenomena

Checkerboard codes, especially in the fracton regime, host localized (“fracton”) excitations with constrained mobility determined by subsystem symmetries. For the U(1) model, isolated star excitations drag semi-infinite constrained diagonals, rendering individual fractons immobile and energetically confined; four-fracton composites regain unrestricted mobility in two dimensions, while lineon excitations along plaquette strings appear similarly (Vieweg et al., 30 May 2025). The XY checkerboard variant supports both Z₂ anyonic excitations (subject to mobility restrictions from diagonal parities) and type-I fracton excitations, which are immobile unless aggregated into composite configurations that move along lines or rectangles (Vieweg et al., 2024). Both types exhibit dimensional reduction of quasiparticle motion, a hallmark of fracton order.

5. Phase Diagram and Orders

Checkerboard toric codes realize multiple quantum phases, depending on anisotropy and flavor parameters:

• Z₂ topological order (toric code-like) for pure x or y star interactions,
• Type-I fracton order in intermediate (anisotropic) regimes, characterized by robust sub-extensive degeneracy and constrained excitations,
• Non-topological, confining phases (U(1) model for all $\lambda > 0$), where subextensive degeneracy is fully lifted and long-range entanglement vanishes (Vieweg et al., 2024, Vieweg et al., 30 May 2025).

Phase transitions—mapped via duality to self-dual Xu-Moore models in the XY case—are strictly first order, corresponding to level crossings in energy and discontinuous subsystem-parity order parameters. At isotropic points, the intermediate fracton phase disappears and a direct first-order transition separates two toric-code–like phases (Vieweg et al., 2024).

6. High-Order Perturbation Theory and Finite-Size Gaps

Degeneracy splitting and gap scaling in checkerboard codes have been analytically resolved using degenerate perturbation theory (Takahashi’s method). In the U(1) checkerboard toric code, the unique ground state is selected at fourth order ($\lambda^4$), with the energy gap scaling linearly with system size for excited diagonal configurations: $$\Delta E = 5.232 \times 10^{-4} L \lambda^4 + \mathcal{O}(\lambda^6).$$ Higher-order expansions yield corrections ($\lambda^6, \lambda^8, ...$), and gap estimates align with quantum Monte Carlo and exact diagonalization results, confirming the lack of topological ground-state degeneracy for $\lambda > 0$ (Vieweg et al., 30 May 2025).

7. Checkerboard Code Capacity and Bounds

The information-theoretic capacity of checkerboard codes on hexagonal lattices is characterized by rigorous upper and lower bounds derived from lattice geometry and constraint structure (Deng et al., 2015). Although detailed derivations are not publicly available, the methodology provably tightens capacity estimates compared to earlier approaches. Establishing capacity bounds is central for evaluating code performance and asymptotic rates in quantum error correction.

---

Recent developments firmly establish checkerboard codes as a framework for exploring quantum order, coding capacity, and emergent phenomena such as fractons and symmetry-enriched topological phases. The interplay of bipartite operator geometry, subsystem symmetries, and duality mappings orchestrates a rich phase structure and excitation landscape, with rigorous analytical and numerical methods confirming key features in thermodynamic and finite-size regimes (Vieweg et al., 30 May 2025, Vieweg et al., 2024, Deng et al., 2015).