Bulk–Boundary Correspondence

• Bulk-boundary correspondence is a principle linking a system’s topological invariants or geometric features with its emergent boundary modes.
• It extends to non-Hermitian, Floquet, and stochastic systems by using generalized Brillouin zones and operator K-theory to accurately predict edge-localized states.
• Modified boundary conditions and spectral formulations provide a robust framework for classifying phase transitions and designing materials with desired interface phenomena.

Bulk-boundary correspondence refers to a set of rigorous relations between topological invariants or geometric features of the bulk (interior) of a physical system and the emergent phenomena at its boundaries or interfaces. Originally developed for quantum Hall and topological insulator systems, the principle underpins a vast array of phenomena, including quantized conductance, protected edge/surface modes, and localized steady states in non-equilibrium and classical systems. Recent research extends bulk-boundary correspondence far beyond static, Hermitian, clean, and quantum systems, revealing its ubiquity in non-Hermitian, disordered, stochastic, and even soft-matter or biological settings.

1. Formal Statement and Traditional Frameworks

In the canonical case of a crystalline, Hermitian topological insulator, the bulk-boundary correspondence asserts that a quantized topological invariant—such as the Chern number,

$$C = \frac{1}{2\pi}\int_{\mathrm{BZ}} \mathrm{Tr} F_{xy}(k) \, d^2k$$

(with $F_{xy} = \partial_x A_y - \partial_y A_x$ the Berry curvature)—predetermines the number of robust gapless modes localized at an interface between topologically distinct phases. This relationship is underpinned by index theorems, K-theory arguments, and exact mappings between bulk and boundary algebras arising from Toeplitz (half-space) extensions (Wang et al., 2015, Alldridge et al., 2019, Mathai et al., 2015).

This paradigm is generalized via operator algebraic methods to arbitrary symmetry classes, spatial dimension, and strong disorder, with the bulk and boundary invariants defined as elements in real or complex K-theory ($K_*(\mathbb{A})$ for bulk, $K_{*-1}(\mathbb{A}_\partial)$ for the boundary), connected by an exact “boundary map” induced by a short exact sequence of $C^*$-algebras:

$$0 \to \mathbb{A}_\partial \otimes \mathcal{K} \to \widehat{\mathbb{A}} \to \mathbb{A} \to 0$$

where $\mathcal{K}$ denotes compact operators on the half-line (Alldridge et al., 2019). This mapping ensures that bulk topological properties determine the allowed boundary phenomena, regardless of disorder or specific boundary conditions provided the gap remains open.

2. Extensions to Non-Hermitian, Floquet, and Classical Systems

Bulk-boundary correspondence is challenged and restructured in non-Hermitian, driven, or stochastic systems:

• Non-Hermitian systems: Standard bulk invariants computed on the usual Brillouin zone with periodic boundary conditions typically fail in the presence of non-reciprocity or gain/loss, due to the non-Hermitian skin effect (bulk eigenstates pile up at boundaries), breaking the direct bulk–boundary link (Imura et al., 2020, Takane, 2021, Zhang et al., 2020, Xiao et al., 2019). Restoration of correspondence demands replacement of the Bloch BZ by a generalized Brillouin zone (GBZ) defined by the open-boundary spectrum. The topological invariant (e.g., non-Bloch winding or Chern number) must be computed on this GBZ, and only then does it correctly predict boundary-localized modes, as established for both 1D and 2D non-Hermitian models.
• Discrete-time and Floquet dynamics: For periodically driven/interacting many-body Floquet systems, bulk and edge invariants are constructed from "flows," functionals that capture information or charge transport between regions under unitary evolution (Zhang et al., 2022). The flow formalism unifies the construction of both bulk and edge invariants, including the integer Rudner winding, the GNVW index (quantifying quantum information flow), and interacting many-body charge-pumping invariants, ensuring bulk and edge indices exactly agree.
• Classical stochastic processes: In one-dimensional Markov chains and many-body exclusion processes, the principle persists as a correspondence between a winding number of the (possibly non-Hermitian) stochastic generator (after a non-Bloch deformation) and the number of boundary-localized steady states, even in nonergodic, disordered, and interacting systems. Explicitly, for winding $w$, there are $w$ boundary-localized zero modes in steady state under appropriate boundary conditions (Sawada et al., 2024).

3. Bulk Invariants, Boundary Conditions, and Modified Formulations

The precise realization of the correspondence depends sensitively on the bulk boundary conditions and the definition of the topological invariant:

• Modified boundary conditions in non-Hermitian models: To capture skin effects, modified periodic boundary conditions (MPBC) introduce a scaling factor $b$ in the boundary overlap condition, generalizing the momentum-space description and enabling exact mapping between bulk and open-boundary properties. When the non-Hermitian Chern number $N[\tilde{\gamma}, b(\tilde{\gamma})]$ (placed in correspondence with the edge spectrum) is computed along a trajectory $b(\tilde{\gamma})$ respecting bulk and boundary gap closings, every quantum phase transition in the bulk matches a transition in the existence of edge states (Takane, 2021, Imura et al., 2020, Zhang et al., 2020).
• Operator-theoretic and K-theoretic formalism: In higher-dimensional, disordered, and interacting systems, the bulk and boundary invariants are formulated as classes in operator K-theory, and the bulk-boundary correspondence is realized via connecting maps in six-term exact sequences. These frameworks rigorously guarantee robustness against disorder, interactions, and even incommensurability (Leung et al., 2018, Alldridge et al., 2019).
• Geometric and spectral generalizations: Quantum-geometric quantities such as the Hilbert–Schmidt quantum distance can function as predictors of interface modes even in topologically trivial flat band systems. The quantum distance $d_\mathrm{max}$ between eigenstates at different momenta signals and fully determines the existence and characteristics of interface-bound boundary modes, providing a new geometric bulk-interface correspondence that operates beyond integer-valued topological invariants (Oh et al., 2022).

4. Model Realizations and Phase Diagram Structures

Several canonical models and methodologies have been used to formulate, verify, and extend bulk-boundary correspondence:

System/Class	Bulk Invariant	Boundary Signature(s)
Hermitian Chern/topological insulator	Chern/Z₂ number (Berry curvature)	Chiral edge/mode, spectrum crossing
Non-Hermitian (Chern, skin effect)	Non-Hermitian Chern/winding (GBZ)	Existence of boundary-localized branch
3D Node-line & exceptional-line semimetal	Winding numbers on GBZ	Hopf link of exceptional lines, skin effect
Floquet many-body localized 2D system	Flow-based quantized index	Rational GNVW edge index
1D stochastic Markov/ASEP process	Bulk winding (non-Bloch)	Localized steady states at boundary
Soft matter (LC+aerosil)	Strength of random disorder	Surface fractal dimension
Flat-band topologically trivial system	Maximum quantum distance d_max	Dispersion/decay of interface bound state

The explicit construction of phase diagrams typically involves tracking bulk invariants (e.g., the non-Hermitian Chern number or winding) as a function of tuning parameters and matching the boundaries with explicit gap closings in the open spectrum. In non-Hermitian higher dimensions, phase boundaries separating distinct topological phases (e.g., $N=0,1$ regions) coincide for both modified bulk and open-boundary geometries, provided bulk and boundary gap closings are correctly incorporated (Takane, 2021).

5. Beyond Quantum and Electronic Systems: Universality and Limitations

Bulk-boundary correspondence exhibits remarkable universality:

• Photonic and Maxwell systems: The correspondence extends to Maxwell waves in photonic crystals; here, the existence of interface modes is tied to gap Chern numbers, which are robustly encoded in the block structure of the Dirichlet-to-Neumann operator for the boundary problem in 2D (Tsukerman, 2024). The generalization from 1D (with scalar impedances) to 2D (infinite-dimensional DtN operators) introduces significant analytical complexity but preserves the topological structure underlying interface states.
• Quasicrystals and aperiodic systems: In Sturmian (quasiperiodic) systems, the only nontrivial topological data are gap-labels (integrated density of states). The winding number of spectral flows under phason cycling matches these bulk gap labels, requiring an augmented or smoothed (circle-valued) phason to formalize the edge invariant (Kellendonk et al., 2017).
• Soft-matter and disordered media: The relation is extendable to surface–bulk correlation of physical roughness exponents: in liquid crystal gels with embedded disorder, the bulk density of disorder is in one-to-one correspondence with the fractal dimension of the surface topography, measurable by atomic-force microscopy (Ramazanoglu et al., 2019).

However, certain scenarios reveal limitations or nontrivial refinements:

• In generic non-Hermitian systems with uncontrolled boundary conditions or disorder, the skin effect and exceptional-point topology require a precise identification of the correct GBZ/invariant for restoration of BBC.
• For fragile topology, twisted boundary conditions (TBCs) selectively close bulk gaps and enable the bulk-boundary correspondence to detect and characterize fragile topological phases via real-space invariants (Song et al., 2019). The number of TBC-induced level crossings is fixed by the associated local quantum numbers.

6. Unified and Atypical Bulk-Boundary Correspondences

Advanced frameworks achieve further generality:

• A “unified” bulk-boundary correspondence for band insulators yields a single formula, in terms of the trace of the bulk Green’s function (or the zeros of a “pole-determinant”), for the exact number of in-gap edge states—valid in all symmetry classes, in the presence of disorder, and even for boundary-junction/geometry-dependent scenarios (Rhim et al., 2017).
• In higher-order topological insulators and crystalline phases (with e.g., mirror, inversion, or rotation), a hierarchy of K-theory subgroups systematically classifies possible codimension-$n$ boundary states via quotient groups $\mathcal{D}^{(n)} = K^{(n-1)}/K^{(n)}$, with each level in the sequence corresponding to edge, hinge, or corner states (Trifunovic et al., 2018).
• Spectral bulk-boundary correspondence (SBBC) extends the principle to energy-dependent Green's functions, allowing the relation to hold in the absence of a spectral gap and in the presence of chiral symmetry breaking and dynamical self-energies, provided an appropriate operator splitting exists (Tamura et al., 2021).

7. Outlook and Open Problems

Research continues on rigorous characterization and extension of bulk-boundary correspondence:

• Analytically complete proofs of correspondence in 2D Maxwell/photonic systems (operator-theoretic approach to infinite-dimensional DtN maps) remain open (Tsukerman, 2024).
• Generalization to systems with non-conserving boundary processes, higher dimensions, many-body and interacting classical and quantum systems, and biologically relevant networks is ongoing (Sawada et al., 2024).
• The role of geometric (non-topological) bulk quantities—such as quantum distance—and their predictive power for interface modes signal new geometric bulk-boundary correspondences, hinting at possible underlying unifying mathematical frameworks beyond topology (Oh et al., 2022).
• The interplay with T-duality (Mathai et al., 2015) and twisted (e.g., real-space) boundary maps (Song et al., 2019) further clarifies and trivializes bulk-boundary correspondences, pointing to deep categorical and homological foundations for the principle.

In sum, bulk-boundary correspondence is a mathematically rigorous, physically robust principle that connects internal (bulk) invariants or order parameters—topological, geometric, or even statistical—to emergent boundary or interface phenomena, guiding both fundamental understanding and materials design across an increasingly broad landscape of physical systems.