Chiral-Unitary Symmetry Classes (AIII)

• Chiral-unitary symmetry classes (AIII) are defined by a unitary chiral operator that anticommutes with the Hamiltonian, resulting in block off-diagonal forms and symmetric energy spectra.
• They exhibit distinctive spectral properties such as universal level repulsion and topologically protected zero-energy modes, as demonstrated in chiral Gaussian unitary ensembles.
• These classes support integer-valued topological invariants in odd dimensions and reveal complex phase transitions, including Anderson localization and multifractal behavior under disorder.

Chiral-unitary symmetry classes, designated AIII in the Altland–Zirnbauer (AZ) taxonomy, comprise complex fermionic Hamiltonians constrained such that a unitary chiral symmetry operator anticommutes with the Hamiltonian. These classes are characterized by the absence of time-reversal and particle-hole symmetries, leading to structures and critical phenomena distinct from standard Wigner–Dyson ensembles. Chiral symmetry enforces block off-diagonal Hamiltonian forms, drives fundamental spectral repulsion near zero energy, permits topological invariants in odd spatial dimensions, and produces novel, higher-order, and anisotropic localizations.

1. Defining Symmetry Constraints and Classification

Class AIII is defined by the existence of a unitary chiral (sublattice) operator $S$ such that

$S\,H\,S^{-1} = -H,\qquad S^2 = 1$

with $H$ a Hermitian operator acting on a complex Hilbert space. In basis where $S = \operatorname{diag}(1_M, -1_N)$ (M, N are the numbers of A- and B-sublattice sites), $H$ takes the canonical off-diagonal form: $$H(k) = \begin{pmatrix} 0 & Q(k) \ Q^\dagger(k) & 0 \end{pmatrix}$$ where $Q(k)$ is a complex $M \times N$ matrix. This structure underlies random-matrix realizations (Chiral-Gaussian Unitary Ensemble, chGUE), physical tight-binding models, and field-theoretic representations (Zirnbauer, 2010, Abramovici et al., 2011, Rehemanjiang et al., 2019).

Within the AZ scheme, chiral symmetry classes include AIII (chiral-unitary, no additional antiunitary symmetry), BDI (chiral-orthogonal), and CII (chiral-symplectic). For AIII, the relevant classifying spaces are complex Stiefel manifolds $V_M(\mathbb{C}^N)$ for non-balanced sublattices, and Grassmannian manifolds $U(N)/U(M)\times U(N-M)$ for balanced cases (Dai et al., 2024).

2. Spectral and Random Matrix Properties

Chiral symmetry ensures the spectrum of $H$ is symmetric about zero, and in non-balanced cases (M≠N) yields $|N-M|$ topologically protected zero-energy modes (flat bands) (Zirnbauer, 2010, Rehemanjiang et al., 2019, Dai et al., 2024). The joint probability density of nonzero eigenvalues $\{\lambda_j\}$ in the chGUE is: $$P(\{\lambda_j\}) \propto \prod_{i<j} |\lambda_i^2-\lambda_j^2|^2 \prod_{k=1}^{n} \lambda_k^{2|M-N|+1} e^{-\frac{1}{2} \sum_k \lambda_k^2}$$ implying universal "hard-edge" level repulsion $\rho(E) \sim |E|^{2|M-N|+1}$ near $E=0$. In balanced cases (M=N), $\rho(E) \sim |E|$, confirmed both theoretically and experimentally (Rehemanjiang et al., 2019). Bulk spectral statistics are Wigner-Dyson (GUE), while near $E=0$, correlation functions are governed by the Bessel kernel (Zirnbauer, 2010).

3. Topological Invariants and Physical Realizations

Chiral-unitary systems admit integer-valued topological invariants (winding numbers) in odd spatial dimensions. In 1D, for a Bloch Hamiltonian $H(k)$ with off-diagonal $\Delta(k)$, the invariant is

$$\nu = \frac{1}{2\pi} \int_{-\pi}^{\pi} d[\arg \det \Delta(k)]$$

This classifies phases such as deformed SSH chains and their couplings to other chiral models (Matveeva et al., 2022).

In 3D, the invariant is a third-homotopy winding number,

$$\nu_{3D} = \frac{1}{24\pi^2} \int_{\text{BZ}} d^3k\, \epsilon^{\mu\nu\rho} \, \mathrm{Tr}\left[ (q^{-1}\partial_\mu q)(q^{-1}\partial_\nu q)(q^{-1}\partial_\rho q) \right]$$

with $q(\mathbf{k})$ the off-diagonal block, or more generally using the chiral operator $S$ (Liu et al., 2023, Dai et al., 2024). For models parametrized as stacks of Lieb or dice lattices, flat zero modes coexist with gapped topological phases (Dai et al., 2024). These invariants govern the existence of protected boundary states, including surface Dirac cones and higher-order (corner) states (Benalcazar et al., 2021).

4. Anderson Transition and Multifractality

Disordered AIII models exhibit Anderson localization phenomena, which differ qualitatively from standard unitary classes due to the enforced chiral symmetry (Wang et al., 2021, Xiao et al., 2022). In 2D, the nonlinear sigma model (NLSM) for class AIII is

$$S[Q] = -\int d^2r\, \Big[ \frac{\sigma}{8\pi}\, \mathrm{Tr}(Q^{-1}\nabla Q)^2 + \frac{c}{8\pi}(\mathrm{Tr} Q^{-1}\nabla Q)^2 \Big]$$

with the additional "Gade" term from the U(1) symmetry (König et al., 2012, Karcher et al., 2023). The metallic phase persists in the absence of topological defects; nonperturbative vortex-antivortex excitations induce a BKT-like metal-insulator transition with critical stiffness $K_c=2$. The field theory and numerics reveal a “generalized multifractality” structure, with scaling exponents labeled by two multi-indices $\lambda,\lambda'$ associated with sublattices, organized by Weyl symmetries (Karcher et al., 2023).

In 3D, the Anderson transition is controlled by proliferation of vortex loops in the NLSM, leading to critical exponent $\nu_{\rm AIII}=1.06\pm 0.02$, which is distinct from the standard unitary class ($\nu_A \approx 1.44$) (Wang et al., 2021, Zhao et al., 26 Jun 2025). Weak topological indices can drive an emergent quasi-localized phase, with delocalization along the topologically nontrivial direction but localization transversely (Xiao et al., 2022, Zhao et al., 2024, Zhao et al., 26 Jun 2025).

Regime	AIII Exponent $\nu$	Phase Structure
3D (AIII, no weak)	$1.06\pm0.02$	Metal $\to$ Insulator
3D (AIII, weak index)	$\nu_M\approx 0.82$, $\nu_Q\approx 1.00$	Metal $\to$ Quasi-localized $\to$ Insulator

The phase diagram for disordered chiral-unitary systems matches that of 3D type-II superconductors: a normal (metallic), mixed (quasi-localized), and insulating (Meissner) phase hierarchy (Zhao et al., 26 Jun 2025).

5. Topological Classification, Stiefel Manifolds, and Boundary Modes

For bipartite lattices with non-equal sublattices (M≠N), gapped AIII Hamiltonians are classified by complex Stiefel manifolds $V_M(\mathbb{C}^N)$: $V_M(\mathbb{C}^N) = U(N)/U(N-M)$ The homotopy groups $\pi_d[V_M(\mathbb{C}^N)]$ classify phases: in 3D, both balanced (M=N) and minimal-imbalance (N−M=1) support integer invariants, directly tied to the count of zero modes (Dai et al., 2024). Systems such as Lieb and dice lattices exemplify these classifications, yielding exactly $|N-M|$ flat bands at zero energy, highly susceptible to correlation-induced physics (Dai et al., 2024).

Boundary modes in strong AIII phases correspond to the values of the topological invariant; in higher-order (multipole) phases, boundary obstruction can host localized corner modes, as revealed by the quantized “multipole chiral number” $N_{xy}$ or $N_{xyz}$ (Benalcazar et al., 2021).

6. Critical Phenomena, Phase Transitions, and Physical Realizations

Metal-insulator transitions in the AIII class arise from the unbinding of vortex-antivortex pairs in 2D, and vortex loop condensation in 3D (König et al., 2012, Zhao et al., 26 Jun 2025). In presence of weak topological terms (Berry-phase contributions from winding numbers), a new quasi-localized regime emerges in which conductance is metallic along the topological direction and exponentially suppressed otherwise (Xiao et al., 2022, Zhao et al., 2024). The phase transitions—metal to quasi-localized to insulator—are determined by both disorder and topological indices, and associated critical exponents differ from conventional universality classes (Wang et al., 2021, Xiao et al., 2022).

Physical realizations include engineered lattice models (deformed SSH chains, Lieb/dice lattice stacks), microwave resonator experiments, and rhombohedral graphite multilayers. In all cases, chiral symmetry is enforced via sublattice engineering and time-reversal breaking, producing characteristic spectral repulsion and topologically robust zero modes (Ho et al., 2017, Rehemanjiang et al., 2019, Liu et al., 2023).

7. Connections, Extensions, and Higher-Order Topology

Chiral-unitary symmetry is the minimal setting admitting block off-diagonal Hamiltonians with integer topological invariants in odd dimensions. Couplings of AIII blocks with their time-reversal partner yield classes DIII and CI, with topological equivalence in $\mathbb{Z}$-valued sectors (Liu et al., 2023, Matveeva et al., 2022). Adiabatic connectivity between these classes depends solely on the preserved chiral operator. Higher-order boundary phenomena, such as robust corner modes, are classified by generalizations of winding numbers to sublattice multipole moments (Benalcazar et al., 2021).

In presence of nontrivial weak indices, chiral-unitary systems exhibit emergent universality classes, anisotropic transport, and a direct analogy to vortex physics in superconductors—establishing deep links between symmetry-based classification, topological defects, and criticality in quantum matter (Xiao et al., 2022, Zhao et al., 26 Jun 2025).

---

References

• (Zirnbauer, 2010) Symmetry Classes
• (Abramovici et al., 2011) Clifford modules and symmetries of topological insulators
• (König et al., 2012) Metal-insulator transition in 2D random fermion systems of chiral symmetry classes
• (Markos et al., 2012) Disordered two-dimensional electron systems with chiral symmetry
• (Ho et al., 2017) Chiral symmetry classes and Dirac nodal lines in three-dimensional layered systems
• (Rehemanjiang et al., 2019) A microwave realization of the chiral orthogonal, unitary, and symplectic ensembles
• (Wang et al., 2021) Universality classes of the Anderson transition in three-dimensional symmetry classes AIII, BDI, C, D and CI
• (Benalcazar et al., 2021) Chiral-Symmetric Higher-Order Topological Phases of Matter
• (Matveeva et al., 2022) One-dimensional non-interacting topological insulators with chiral symmetry
• (Xiao et al., 2022) Anisotropic Topological Anderson Transitions in Chiral Symmetry Classes
• (Karcher et al., 2023) Generalized multifractality in 2D disordered systems of chiral symmetry classes
• (Liu et al., 2023) Elementary models of 3D topological insulators with chiral symmetry
• (Zhao et al., 2024) Topological effect on the Anderson transition in chiral symmetry classes
• (Dai et al., 2024) Topological classification for chiral symmetry with non-equal sublattices
• (Zhao et al., 26 Jun 2025) Theory of the Anderson transition in three-dimensional chiral symmetry classes: Connection to type-II superconductors