Topological characterization of multifold band degeneracies in Altland-Zirnbauer symmetry classes

Technical Summary: Topological Characterization of Multifold Band Degeneracies in Altland-Zirnbauer Symmetry Classes

Problem Statement
Topological band theory traditionally characterizes band degeneracies (nodal points, lines, or surfaces) using invariants defined on an enclosing sphere where the energy spectrum remains gapped. This "enclosing sphere paradigm" has been successfully completed for minimal degeneracies (typically two-fold) across all ten Altland-Zirnbauer (AZ) symmetry classes. However, higher-order degeneracies ($n$-fold, where $n > 2$) have historically been studied almost exclusively under the protection of crystalline symmetries, which confine these nodes to high-symmetry points or lines.

This work addresses the complementary question: How are generic $n$-fold band degeneracies characterized and protected when their stability derives solely from local AZ symmetries, without invoking crystalline symmetry? A fundamental obstacle arises in extending the standard paradigm: an $n$-fold node is the intersection of two loci supporting $(n-1)$-fold degeneracies. These loci inevitably pierce any enclosing sphere, preventing the existence of a uniform spectral gap. Consequently, the standard homotopy classification based on maps from the enclosing sphere to a single-gap classifying space fails.

Methodology
The authors develop a general strategy to overcome the lack of a uniform gap by turning the obstruction itself into the diagnostic tool. The methodology proceeds as follows:

1. Codimension Analysis: Using von Neumann-Wigner counting arguments adapted for the ten AZ classes, the authors derive the codimension $\delta_{CL}(n)$ required to form a generic $n$-fold degeneracy. They find that for Hermitian AZ classes, this codimension grows quadratically with $n$ (e.g., $\delta \sim n^2$). This implies that generic multifold nodes are not accessible in standard 3D momentum spaces but require higher-dimensional parameter spaces (combining momenta with tuning parameters or synthetic dimensions).
2. Nodal Manifolds: Instead of a gapped sphere, the authors define two "nodal manifolds," $M_1$ and $M_2$, as the intersections of the $(n-1)$-fold degeneracy loci ($L_1$ and $L_2$) with the enclosing sphere $S^\ell$ (where $\ell = \delta - 1$). On each manifold, a complementary spectral gap is restored (e.g., on $M_1$, the gap between the lower $n-1$ bands and the upper band is open).
3. Linking and Invariants: The core insight is that the topological stability of the $n$-fold node is equivalent to the robust linking of $M_1$ and $M_2$ within the enclosing sphere. The authors establish a two-way correspondence:
   • If $M_1$ and $M_2$ are robustly linked, the $n$-fold node is topologically protected.
   • The topological band invariants (Chern numbers, Stiefel-Whitney numbers, winding numbers) defined on nontrivial cycles of one nodal manifold encode the linking numbers with cycles of the other manifold.
4. Mathematical Formalism: The relationship between band invariants and linking is formalized using differential forms and cohomology. The authors demonstrate that band invariants on a submanifold $U \subset M_1$ can be expressed as integrals of differential forms that effectively count the intersection of $U$ with a dual manifold in $M_2$, thereby identifying the invariant as a linking number. This framework accommodates both integer ($\mathbb{Z}$) and $\mathbb{Z}_2$ invariants.
5. Case-by-Case Analysis: The authors apply this framework to minimal $n$-band models for all ten AZ classes. They explicitly identify the geometry of the nodal manifolds (often identifying them with standard classifying spaces like $\mathbb{R}P^2$, $\mathbb{C}P^2$, $\mathbb{H}P^2$, or unit tangent bundles) and compute the relevant band invariants on their nontrivial cycles.

Key Contributions and Results

• Codimension Scaling: The paper provides a general formula for the codimension of $n$-fold degeneracies in all AZ classes. For example, in Class AI, $\delta(n) = \frac{1}{2}(n-1)(n+2)$; in Class A, $\delta(n) = n^2 - 1$. This confirms that generic multifold nodes require parameter spaces of dimension $\delta(n)$.
• Reformulation of Topological Protection: The work recasts the topology of multifold nodes as the topology of linked nodal manifolds. It proves that the stability of the node is not due to a single global invariant on a gapped sphere, but rather the non-trivial linking of the loci where gaps close.
• Explicit Characterization in Minimal Models:
  • Class AI (Spinless TRS): For a 3-fold node in 5D, the nodal manifolds are identified as $\mathbb{R}P^2$. The linking is detected by non-trivial first and second Stiefel-Whitney numbers ($w_1, w_2$) on 1-cells and 2-cells.
  • Class A (Unitary): For a 3-fold node in 8D, the manifolds are $\mathbb{C}P^2$. Linking is characterized by first and second Chern numbers ($C_1, C_2$).
  • Class AII (Spinful TRS): For a 3-fold node (6-fold physical degeneracy) in 13D, manifolds are $\mathbb{H}P^2$. Linking is characterized by second and fourth Chern numbers.
  • Chiral Classes (BDI, AIII, CII): The analysis distinguishes between nodal manifolds at zero energy ($M_\gamma$) and non-zero energy ($M_\emptyset$). For Class BDI, the manifolds are a torus ($S^1 \times S^1$) and a disjoint union of circles ($S^1 \sqcup S^1$), linked via Pfaffian signs and winding numbers. For Class AIII, the manifolds involve unit tangent bundles and $U(2)$, linked via Chern numbers and higher-dimensional winding numbers.
  • BdG Classes (D, C, CI, DIII): Similar linking phenomena are established, often involving Pfaffian signs and specific homology generators.
• Generalization of Known Phenomena: The framework reproduces established results for minimal degeneracies (e.g., the linking of nodal rings in 3D Class AI) as low-dimensional corollaries, validating the approach.

Significance and Claims
The authors claim that this work provides a foundation for a general characterization of multifold band degeneracies in models with arbitrarily many bands, moving beyond the limitations of minimal models and crystalline symmetry protection. By establishing a correspondence between band invariants and linking numbers, the paper offers a diagnostic tool that remains applicable even when the standard enclosing sphere is not gapped.

The significance lies in:

1. Completing the AZ Classification: It extends the topological classification of band nodes from minimal degeneracies to generic $n$-fold degeneracies purely within the AZ symmetry framework.
2. Unifying Homotopy and Cohomology: It demonstrates that cohomological tools (linking numbers, Stiefel-Whitney classes) are indispensable for characterizing multifold nodes, complementing the traditional homotopy group approach.
3. Physical Relevance: While generic multifold nodes require high-dimensional parameter spaces, the paper notes that crystalline realizations can be viewed as symmetry-constrained slices of this universal parameter space. The results thus provide a deeper theoretical understanding of why certain crystalline symmetries enforce specific degeneracies and how they might be tuned or probed in synthetic dimensions (e.g., cold atoms, photonics).

The authors remain modest regarding the stability of these invariants in the presence of non-Abelian braiding in multi-gap systems, noting that while the linking picture is robust, the reduction of integer invariants to $\mathbb{Z}_2$ in specific contexts (like Class AI with many bands) requires further investigation rooted in homological constructions. They also leave the extension to multifold nodal lines and surfaces, and formulations in terms of cobordism invariants, as future directions.