The classification of topological phases of matter represents a major milestone in modern condensed matter physics. Early frameworks succeeded by grouping phases according to simple, global, internal symmetries—such as time-reversal or particle-hole symmetry—acting uniformly across a system. These are known as on-site symmetries because their operations can be applied to each lattice site independently without regard to spatial geometry.
However, real-world materials are rarely governed solely by on-site symmetries. They exhibit rich crystalline structures, complex magnetic orderings, non-Hermitian energy dissipation, and time-dependent driving forces. To capture these phenomena, researchers are actively generalizing topological classification frameworks to broader, more intricate symmetry classes. This shift requires moving past traditional mathematical toolkits to understand how unconventional symmetries constrain, protect, and enrich quantum matter.
The On-Site Paradigm and Its Boundaries
To understand why generalizing symmetries is necessary, we must first look at how symmetry protects topology. In a one-dimensional quantum spin chain, certain phases cannot be smoothly deformed into a trivial state without crossing a phase transition, provided specific symmetries are maintained.
Frank Pollmann et al. demonstrated this by showing that the classic Haldane phase of spin-1 chains is uniquely characterized by a universal double degeneracy across its entire entanglement spectrum. This degeneracy is robustly protected by inversion or time-reversal symmetries, establishing a clear, wavefunction-based framework for identifying symmetry-protected topological (SPT) phases in one dimension.
Standard On-Site Symmetry:
[Site 1] ──> [Site 2] ──> [Site 3] (Symmetry acts locally on each individual site)
Spatial/Crystalline Symmetry:
[Site 1] ───────────────> [Site 3] (Symmetry maps positions across space, e.g., Inversion)
When moving to two dimensions and introducing strong electron-electron interactions, the mathematical landscape becomes significantly more complex. For interacting fermionic systems, Meng Cheng et al. expanded the classification of 2D SPT phases by analyzing the topological properties of symmetry defects—specifically how these defects fuse and braid around one another. While this approach successfully classifies phases governed by finite, on-site unitary groups, the authors noted that extending these models to continuous groups or magnetic space groups requires deeper theoretical refinement to handle the spatial constraints.
The Interplay of Topology and Global Symmetries
When a system possesses both intrinsic topological order (like fractionalized anyonic excitations) and global symmetries, it forms a Symmetry-Enriched Topological (SET) phase. In these systems, the global symmetry can be "fractionalized" by the anyons, meaning an individual anyon might carry a fraction of a symmetry charge.
A persistent challenge has been rigorously modeling this interplay within a single microscopic framework. To address this, Liang Chang et al. extended the exactly solvable Levin-Wen model from unitary fusion categories to unitary multi-fusion categories. This mathematical expansion provides a controlled playground for studying how global symmetries interact with intrinsic topological order at the microscopic level.
Despite these mathematical advances, translating abstract classifications into concrete field theories remains difficult. Zhen Bi et al. utilized semiclassical nonlinear sigma models to bridge this gap for bosonic phases. However, the authors emphasized that a complete classification that fully resolves the complex feedback loop between fractionalized excitations and global symmetries in SET phases remains an ongoing research objective.
Moving Beyond On-Site: Crystalline and Spatial Symmetries
Crystalline symmetries—such as rotations, inversions, and reflections—fundamentally differ from on-site symmetries because they map one spatial coordinate to another. This non-local action introduces geometric constraints that alter the underlying topological invariants.
Consider the Hopf insulator, a three-dimensional topological phase characterized by a unique linking number of its energy bands in momentum space. Bastien Lapierre et al. generalized this concept to multi-band ($N$-band) systems, establishing a bulk-boundary correspondence where the bulk's quantized magnetoelectric response manifests as a quantized Hall conductivity on the boundary. Crucially, the authors noted that explicitly accounting for crystalline symmetries within this multi-band framework is a vital next step for comprehensive classification.
When rotation symmetry is explicitly integrated into these models, it can uncover entirely new topological sub-classes. Aleksandra Nelson et al. introduced a novel topological invariant called the "returning Thouless pump." By combining this invariant with rotation symmetry, the authors revealed "delicate" topological phases that possess unique surface properties and escape standard Hopf classifications.
Standard Topology: Well-defined, localized Wannier functions possible.
Chern Insulators: Localized Wannier functions mathematically impossible.
Weak Crystalline Topology: Inversion symmetry restores quantization of bound charges/currents.
Spatial symmetries also serve as diagnostic tools when standard methods fail. In Chern insulators, the non-zero topology makes it mathematically impossible to construct localized real-space wavefunctions (Wannier functions) to define electric polarization. To circumvent this hurdle, Mikael C. Rechtsman et al. demonstrated that despite the lack of Wannier functions, physical manifestations of polarization—such as bound charges and currents—remain well-defined and can be quantized by inversion symmetry, pointing toward a broader role for crystalline symmetries in weak topological phases.
Similarly, in gapless systems, Yuting Qian et al. leveraged the mathematical "mismatch" between symmetry indicators under $S_4$ crystal symmetry to identify hidden Weyl semimetals, highlighting the ongoing challenge of developing universal symmetry indicators across all noncentrosymmetric space groups.
Higher-Order Boundaries and Real-Space Obstructions
Conventional topological materials obey a strict bulk-boundary correspondence: a $d$-dimensional bulk hosts gapless modes on its $(d-1)$-dimensional boundary (e.g., a 3D topological insulator has 2D gapless surface states). Higher-Order Topological Insulators (HOTIs) break this rule. Their surfaces are completely gapped, but they host protected, gapless modes at lower-dimensional boundaries, such as 1D hinges or 0D corners.
Yizhi You et al. developed theoretical frameworks to understand these higher-order states in both interacting bosonic and fermionic systems. They demonstrated that these states feature a much richer bulk-boundary correspondence, though a complete classification across all possible symmetry groups and dimensions remains an open question.
Conventional Boundary: 3D Bulk ──> 2D Gapless Surface
Higher-Order Boundary: 3D Bulk ──> 2D Gapped Surface ──> 1D Gapless Hinge / 0D Corner
To understand why these states remain topologically protected despite having gapped surfaces, consider the concept of Surface Topological Order (STO). By introducing interacting quantum states to the boundary, Apoorv Tiwari et al. showed how to "unhinge" the surface and hinge states of HOTIs. This mechanism allows the boundary to be fully gapped while preserving the crucial global symmetries—specifically $C_{2n}T$ symmetry—that protect the underlying bulk topology.
From a real-space perspective, these higher-order phases can be diagnosed by looking at the fundamental units of the system. In topological superconductors, Frank Schindler et al. analyzed the system using real-space Cooper pair wavefunctions. They discovered that in the topological phase, Cooper pairs cannot be perfectly localized in real space while maintaining the protecting symmetries, creating a "pairing obstruction." This perspective provides a powerful tool for diagnosing second-order topological superconductors, though the current framework is limited to systems that lack time-reversal symmetry.
Unconventional Symmetry Domains: Non-Hermitian, Altermagnetic, and Spatiotemporal
The modern frontier of this research direction extends the classification of topology into entirely unconventional symmetry domains:
- Non-Hermitian Systems: When systems experience energy gain or loss, their Hamiltonians become non-Hermitian, and their energy eigenvalues become complex numbers. M. Michael Denner et al. discovered the Exceptional Topological Insulator (ETI) within non-Hermitian systems. The ETI features unique surface states that are mathematically impossible in standalone 2D systems, though how explicit symmetry constraints might further refine or sub-classify these non-Hermitian phases remains unknown.
- Altermagnetism: Altermagnets represent a recently highlighted magnetic class where opposite spin sublattices are connected by spatial rotations rather than translations or inversion. Shuai Qu et al. demonstrated that unique spin symmetries protect Weyl node-network metals in altermagnetic materials, opening up a broad, unexplored territory for spin-symmetry-protected topology.
- Spatiotemporal Symmetries: Symmetries can also be extended into the time domain. In periodically driven (Floquet) quantum systems, Gilad Kishony et al. proposed "time vortices"—novel space-time defects that act as traps for unique $\pi$ Majorana modes. This framework currently operates under particle-hole symmetry, leaving the extension to other time-dependent symmetry classes as an open challenge.
- Higher-Form Symmetries: Instead of acting on local objects (like sites), higher-form symmetries act on extended objects like lines or surfaces. Dan Mao et al. explored how an emergent $U(1) \times U(1)$ 1-form symmetry governs fractionalization in twisted bilayer graphene, allowing for fractional charges and "fractonic" excitations with restricted mobility without breaking time-reversal symmetry.
By systematically dismantling the restrictions of simple on-site symmetries, this collective body of research is assembling a unified framework capable of classifying topological matter across the full diversity of physical symmetries found in nature.
Key Foundational Work
Classification and description of bosonic symmetry protected topological phases with semiclassical nonlinear sigma models (2015) — contributed 2 concept elements to this direction.
On enriching the Levin–Wen model with symmetry (2015) — contributed 2 concept elements to this direction.
Entanglement spectrum of a topological phase in one dimension (2010) — contributed 1 concept elements to this direction.