Spectral Aharonov–Bohm Effect

• Spectral Aharonov–Bohm effect is a quantum phenomenon where the energy spectrum of Hamiltonians depends on a global magnetic flux despite a locally vanishing magnetic field.
• It manifests through distinct flux-dependent modifications such as shifts in eigenvalue distributions, spectral flows, and singularities in the wave trace, revealed using microlocal and algebraic methods.
• The effect extends across various geometries—including planar annuli, tori, and almost-Riemannian spaces—offering experimentally testable predictions via observable spectral shifts.

The spectral Aharonov–Bohm (AB) effect describes the dependence of the energy spectrum of quantum Hamiltonians on the global holonomy (flux) of a vector potential, even when the corresponding field strength vanishes identically in the physical region. This phenomenon is a spectral manifestation of the original AB effect, in which quantum phases acquired by wavefunctions around noncontractible loops translate into observable shifts in spectra. Unlike the classical AB setup emphasizing interference, the spectral effect uncovers flux-dependent modifications of eigenvalue distributions, spectral flows, and wave trace singularities in multiply connected, topological, or otherwise flux-sensitive quantum systems.

1. Mathematical Framework and Operator Definitions

The prototypical context for the spectral AB effect is the magnetic Schrödinger operator on domains with nontrivial topology. Consider, for instance, the planar annulus $Q_R = \{ x \in \mathbb{R}^2 : 1 < |x| < R \}$, with Dirichlet boundary conditions. The Hamiltonian is

$H_{A,V}\;=\; (i\nabla + A(x))^2 + V(x)$

with $A$ and $V$ smooth, and with $B(x) = \nabla \times A = 0$ in $Q_R$. Although $A$ is locally a pure gauge ($A = \nabla \chi$), globally it may have nontrivial holonomy due to the topology of $Q_R$, allowing for a nonzero magnetic flux

$\Phi = \int_\gamma A \cdot dx$

for closed curves $\gamma$ winding around the inner boundary. The normalized holonomy is $\alpha = \Phi/2\pi$.

In many relevant variants—punctured planes, tori, or more singular geometries (such as almost-Riemannian manifolds)—the central feature remains: the magnetic potential is curl-free in the accessible region, but the topology enforces gauge-invariant, physically detectable effects through the spectrum (Eskin et al., 2013, Boscain et al., 2014).

2. Spectral Consequences: Wave Trace Singularities and Flux Dependence

The spectral AB effect is manifested in precise asymptotic formulas, notably via singularities in the wave trace: $$T(t) = \sum_j \cos(t\sqrt{\lambda_j}) = \int_{\Omega} E(x,x,t)\,dx$$ where $\{\lambda_j\}$ are eigenvalues of $H_{A,V}$.

In the planar annulus with the geometry chosen so that the minimal periodic broken geodesics are equilateral triangles, $T(t)$ has a singularity at $L = 3R\sqrt{3}$. The leading singularity is

$\Sing_{t=L}\;T(t)\;=\; -\,2^{-\tfrac52} 3^{\tfrac14} R^{\tfrac32}\cos(\Phi) (t-3R\sqrt{3})_+^{-3/2}$

demonstrating explicit dependence on the flux $\Phi$ via a $\cos(\Phi)$ prefactor. This establishes that, despite the vanishing magnetic field, the spectrum is a function of the enclosed flux: a direct signature of the AB effect in spectral data (Eskin et al., 2013). The proof relies on microlocal analysis of wave group propagators along periodic orbits, incorporating the phase acquired along noncontractible loops.

On the flat torus $\mathbb{R}^2/L$ ($L$ a rank-2 lattice), $A$ is gauge-equivalent to a constant vector $A_0$, and the two independent fluxes $\Phi_j = \langle A_0, e_j \rangle$ control spectral features: $\Sing_{t=|d|} T(t) \propto \cos(A_0 \cdot d)$ for $d \in L$. Thus, the spectrum detects the vector of holonomies modulo $2\pi$.

A similar dependence is seen in strictly two-dimensional, radially symmetric domains with confining potentials and impenetrable solenoids. In such settings, the eigenvalues of the self-adjoint extension,

$$H(\alpha) = -\partial_r^2 - \frac{1}{r}\partial_r + \frac{(m+\alpha)^2}{r^2} + V(r)$$

display $1$-periodic, even-in-$\alpha$ dependence with minima at integer $\alpha$ and maxima at half-integers, illustrating that the ground state energy oscillates with the enclosed flux (Oliveira et al., 2017).

3. Manifolds and Generalizations: Effects on Almost-Riemannian and Singular Geometries

The spectral AB effect persists across a broad class of geometries, including almost-Riemannian manifolds and conic singularities. In the Grushin cylinder $M = \mathbb{R}_x \setminus \{0\} \times S^1_\theta$, and the Grushin sphere, a closed, but not exact, AB potential $A = \alpha\,d\theta$ is introduced, modifying the magnetic Laplace-Beltrami operator: $\Delta_\alpha = (d + A)^*(d + A)$ leading to flux-dependent eigenvalues. On the Grushin cylinder,

$E_{n,k}(\alpha) = 4n |k-\alpha|$

with a transition between purely discrete and continuous spectrum as $\alpha$ varies through integers—the $k$-th Fourier channel becomes continuous precisely when $\alpha \in \mathbb{Z}$. On the sphere, flux modifies the finite degeneracies, with level crossings (degeneracy jumps) occurring at half-integer values (Boscain et al., 2014).

Similar spectral flow and accumulation phenomena arise generically in nontrivial topologies, driven by the singular nature of the volume form and the topology of punctured or multiply connected spaces.

4. Representation-Theoretic Approaches and Quantum Spectroscopy

The AB flux-induced spectral shifts can be understood by a purely algebraic route. For a two-dimensional charged particle in a harmonic trap enclosing a flux $\Phi$, the Hamiltonian $$H = \frac{1}{2M}(p - eA(r))^2 + \frac{1}{2}M\omega^2 r^2$$ exhibits an $o(2,1)$ algebra structure. The energy spectrum is given by

$E_{n,m}(\alpha) = \omega [2n + 1 + |m + \alpha|]$

where $\alpha = e\Phi/(2\pi)$. As $\omega \to 0$ (free particle limit), the standard AB scattering phase shifts emerge, confirming that discrete spectral features encode the AB effect (Hagen, 2012). This demonstrates that the flux modifies angular momentum ladders, producing observable shifts in oscillator levels—a scenario ripe for direct spectroscopic verification.

An explicit proposal involves placing HCl gas in a region with nonzero vector potential $\mathbf{A}$ but vanishing magnetic field (the interior of a toroidal coil). The vibrational energy levels of HCl are shifted according to

$$E_{\pm}(A) = \hbar\omega_0\left[1 \pm \sqrt{\frac{1}{4}+(\alpha/ \hbar\omega_0)^2}\right]$$

where $\alpha$ is proportional to the applied vector potential. The resulting spectroscopic shifts are experimentally accessible and directly track the applied vector potential, providing a local, non-interferometric confirmation of the AB phase (Laganá, 2014).

5. Partial Spectral Flow, Many-Body Extensions, and Topological Phases

The spectral AB effect extends to many-body and topological condensed matter systems. In tubular graphene, adiabatic insertion of an AB flux through a magnetic-field-free region induces spectral flow: energy levels of the tight-binding Hamiltonian cross the Fermi level, creating electron-hole pairs. This is quantified through the notion of partial spectral flow—counting the eigenvalue crossings restricted to subspaces associated with Dirac valleys: $sf_L\{H_t\} = q$ for $q$ flux quanta and suitable subspace $L$ (Katsnelson et al., 2020). Although the total spectral flow vanishes (since the Hamiltonian is unitarily equivalent at $t=0$ and $t=1$), partial spectral flow reveals the physical processes—pair creation—mirroring continuum Dirac predictions.

In bosonic symmetry-protected topological (SPT) phases, twisting the boundary conditions with a $\mathbb{Z}_N$ flux threads a non-onsite symmetry. The edge spectrum is reorganized—the energy levels are shifted according to: $$E_{n,m}(\phi) = \frac{2\pi v}{L} \left[ (n + p/N)^2 / R^2 + R^2(m+1/N)^2 / 4 \right]$$ Realizations in both conformal field theory and explicit lattice regularizations confirm robust spectral responses to the inserted flux, now in an inherently many-body setting (Santos et al., 2013).

6. Generalizations, Open Directions, and Physical Interpretations

Spectral AB effects are robust under deformations: in the analysis of broken geodesics, the dependence on flux $\Phi$ generalizes from equilateral triangles to regular $N$-gons, with corresponding singularities in the wave trace retaining the $\cos(\Phi)$ structure where the closed geodesic orbits are isolated (Eskin et al., 2013). In the torus setting, while cosines of the holonomies are always present, whether the full phase $e^{i\Phi_j}$ can be recovered from the spectrum alone in the presence of an asymmetric potential remains unresolved.

In all settings—Schrödinger operators on domains, Laplacians on singular spaces, tight-binding models on lattices, topological quantum field theories—the key invariant is the nontrivial holonomy of the vector potential, as sensed by quantum states through their spectral data. The spectrum, sensitive only to gauge-invariant combinations, thereby encodes the nonlocality at the heart of the AB effect, manifesting as measurable energy shifts, spectral flows, or reorganizations tied directly to the topology of the underlying space or the global symmetries of the system.

The spectral AB effect thus bridges mathematical spectral theory, microlocal analysis, condensed matter physics, and quantum topology, offering a set of rigorous, experimentally testable consequences of the physical reality of electromagnetic potentials in quantum mechanics, independent of the local presence of field strength.