Anomaly-Induced Phenomena with Massive Fermions: Higher-Landau-Level Dominance from Spatially Modulated Electric Fields

Abstract: We investigate the axial Ward identity for massive fermions under a constant magnetic field at arbitrary strength, maintaining an arbitrary spacetime configuration of a perturbative electric field. We show that a spatially modulated electric field prevents the exact cancellation between the anomaly and pseudoscalar terms, generating a local axial-charge source even in the adiabatic regime where the frequency is subthreshold to massive-fermion production. Remarkably, unlike conventional magnetic responses, this charge generation is dominated not by the contribution of the lowest Landau level, but by those of the higher Landau levels. Our findings provide a microscopic foundation for anomaly-induced transport and anomalous optical responses in gapped systems. In particular, we find that the spatially modulated chiral magnetic effect in weakly gapped Weyl semimetals that exhibits a linear suppression of the magneto-resistance by the magnetic-field strength instead of the renowned quadratic suppression.

• The paper demonstrates that spatially modulated electric fields generate a nonzero local axial-charge source in massive fermions, contradicting the conventional LLL-based intuition and highlighting the role of higher Landau levels.
• The authors calculate axial current divergences in external electromagnetic fields, showing dominance from higher Landan levels with significant precision.
• The findings lead to a new experimental signature of linear magneto-resistance in weakly gapped Weyl semimetals, offering a distinct fingerprint of chiral magnetic effects.

The axial Ward identity (AWI) for massive Dirac fermions in external electromagnetic fields has long been understood to admit a complete cancellation between the anomaly term and the pseudoscalar (PS) mass term whenever both fields are homogeneous and constant, erasing any net axial-charge source even for gapped fermions. In "Anomaly-Induced Phenomena with Massive Fermions: Higher-Landau-Level Dominance from Spatially Modulated Electric Fields" (2608.17611), Hattori, Mameda, Uchiyama, and Yang demonstrate that this washout is not generic: a spatially modulated perturbative electric field superposed on a constant magnetic field produces a nonzero local axial-charge source that survives in the adiabatic regime, where $\omega \ll m$ and real-particle production is kinematically forbidden. The source is dominated by higher Landau levels (hLLs) rather than the lowest Landau level (LLL), overturning the standard LLL-based intuition for anomaly saturation.

Anomaly diagrams resummed over Landau levels

The authors compute the triangle matrix element $iq_\mu \Gamma^{\mu\nu}(q)$ of the diverged axial current coupled to a perturbative photon of four-momentum $q^\mu$, with the fermion propagator replaced by the Schwinger-resummed propagator in a constant magnetic field $B$ at arbitrary strength. The result takes the compact form

$$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$

governed by three dimensionless parameters: $\lambda = q_\parallel^2/m^2$, $\xi_\perp = |q_\perp|^2/(2|eB|)$, and $b = 2|eB|/m^2$. The weight function splits as $w = w_{\rm anom} + w_{\rm PS}$, where $w_{\rm anom}$ is a telescoping double series over Landau-level indices built from the transverse convolution integrals $iq_\mu \Gamma^{\mu\nu}(q)$0 involving Laguerre polynomials.

A central technical achievement is the evaluation of $iq_\mu \Gamma^{\mu\nu}(q)$1 via the generating function

$iq_\mu \Gamma^{\mu\nu}(q)$2

which yields exactly $iq_\mu \Gamma^{\mu\nu}(q)$3 after summing all levels. This establishes that chiral anomaly in $iq_\mu \Gamma^{\mu\nu}(q)$4 dimensions is reproduced only by the full Landau-level sum; the LLL contribution alone is $iq_\mu \Gamma^{\mu\nu}(q)$5, so anomaly saturation is emphatically not "LLL exact" for spatially modulated fields — a direct contradiction of the conventional LLL approximation that underlies much of the chiral-anomaly literature.

Local axial-charge source in the adiabatic limit

In the adiabatic limit $iq_\mu \Gamma^{\mu\nu}(q)$6, the fate of the PS term depends on the hierarchy between $iq_\mu \Gamma^{\mu\nu}(q)$7, $iq_\mu \Gamma^{\mu\nu}(q)$8, and $iq_\mu \Gamma^{\mu\nu}(q)$9. For $q^\mu$0, one recovers the familiar complete cancellation $q^\mu$1. However, for $q^\mu$2, the denominators $q^\mu$3 reduce to $q^\mu$4 only for the LLL pair $q^\mu$5 and to $q^\mu$6 for all other pairs, leaving

$q^\mu$7

which cancels precisely the LLL part of the anomaly but leaves $q^\mu$8. The physical origin is that hLLs become approximate chirality eigenstates when $q^\mu$9: their large cyclotron momentum $B$0 suppresses chirality mixing by $B$1, as the authors verify both from the spinor structure of the PS matrix element and by an explicit Ritus-basis solution of the massive Dirac equation.

Two features distinguish this source. First, it exists entirely within the adiabatic regime, unlike previously known mechanisms relying on Schwinger pair production or above-threshold frequencies. Second, it is strictly local: integrating the AWI over space sets $B$2 and restores $B$3, so no global chirality imbalance accumulates. The charge is instead generated over the magnetic length scale $B$4 and is therefore accessible to probes with finite transverse momentum $B$5. The weak-field hierarchy $B$6 is handled separately via a proper-time representation and expansion in $B$7, recovering Adler's classic result at leading order and confirming momentum-space isotropy.

Anomalous birefringence and spatially modulated CME

Coupling the AWI to Maxwell's equations with a CSE-induced axial current and an axial relaxation time $B$8, the authors derive an effective CME conductivity

$B$9

with $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$0 the axial susceptibility. For photons polarized parallel to $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$1 propagating perpendicular to it, the dispersion relation acquires a mass-dependent correction: for $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$2, the phase velocity becomes $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$3, linear rather than quadratic in $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$4, while the orthogonal polarization remains light-like. This constitutes anomaly-induced birefringence including the fermion mass effect.

For transport, the authors consider a static electric field parallel to $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$5 whose amplitude is modulated in the transverse plane ($$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$6, $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$7). Since $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$8, the adiabatic result applies directly, giving $$iq_\mu \langle k|\tilde j_A^\mu(q)|0\rangle = \frac{e^2}{2\pi^2}\langle k|\tilde{\bm E}(q)|0\rangle\cdot{\bm B}\; w(\lambda,\xi_\perp,b),$$9. For a Gaussian profile $\lambda = q_\parallel^2/m^2$0, the resulting current density along the cylinder axis,

$\lambda = q_\parallel^2/m^2$1

with $\lambda = q_\parallel^2/m^2$2, admits closed-form expressions for the integrated current $\lambda = q_\parallel^2/m^2$3 in both limits. Most consequentially, in the weak-modulation regime $\lambda = q_\parallel^2/m^2$4 the magneto-resistance scales as $\lambda = q_\parallel^2/m^2$5 — linear suppression — whereas the quadratic $\lambda = q_\parallel^2/m^2$6 scaling regarded as the canonical CME signature in uniform fields emerges only for strong modulation $\lambda = q_\parallel^2/m^2$7. This linear-in-$\lambda = q_\parallel^2/m^2$8 magneto-resistance offers a qualitatively new experimental fingerprint for weakly gapped Weyl semimetals, distinct from the negative longitudinal magnetoresistance proposed by Son and Spivak and observed in materials such as Na$\lambda = q_\parallel^2/m^2$9Bi and Cd$\xi_\perp = |q_\perp|^2/(2|eB|)$0As$\xi_\perp = |q_\perp|^2/(2|eB|)$1.

Limitations and open questions

The analysis carries several explicit caveats. The Ohmic conductivity $\xi_\perp = |q_\perp|^2/(2|eB|)$2 is treated as a phenomenological constant rather than derived microscopically; the authors note that a Kubo-formula or kinetic-theory computation of the conductivity remains to be performed. The axial relaxation time $\xi_\perp = |q_\perp|^2/(2|eB|)$3 is likewise introduced phenomenologically, and inter-cone scattering physics enters only through this parameter. The non-adiabatic regime, where real-particle production via the Schwinger mechanism competes with the local source identified here, is deferred to future work, as are finite-temperature and radiative corrections to the PS term. The application to Weyl semimetals assumes ultra-relativistic quasiparticles with $\xi_\perp = |q_\perp|^2/(2|eB|)$4 and relies on equilibrium relations such as $\xi_\perp = |q_\perp|^2/(2|eB|)$5; whether the predicted linear magneto-resistance survives realistic disorder and sample-size effects at the required modulation scale $\xi_\perp = |q_\perp|^2/(2|eB|)$6 is left unaddressed. Finally, the weak-field construction requires performing the full Landau-level summation before expanding, and its convergence condition $\xi_\perp = |q_\perp|^2/(2|eB|)$7 must be relaxed by analytic continuation.

Conclusion

This work establishes that spatial inhomogeneity of a perturbative electric field prevents the anomaly–pseudoscalar cancellation that otherwise eliminates axial-charge production in gapped fermions, yielding a local source sustained by higher-Landau-level contributions even at subthreshold frequencies. The demonstration that anomaly saturation requires the complete Landau-level sum revises the LLL-centric picture underlying chiral magnetic transport, and the resulting prediction of linearly suppressed magneto-resistance in spatially modulated CME provides a concrete, testable signature in gapped topological materials.