Charge-Sign-Dependent Dark Gauge Coupling: A C-Violating Extension of the Dark Photon Framework (final correction)

[CRITICAL REASSESSMENT — 2026-06-02] On adversarial consistency review (gauge invariance, CPT, and anomalies), the central mechanism of this model appears to be forbidden: the charge-sign-dependent coupling requires a non-conserved current (b†b + d†d is not conserved in QED), and CPT pins the matter/antimatter coupling magnitudes equal. The claims of renormalizability and anomaly-freedom are withdrawn; the negative-frequency framing is retired; the model is flagged as an open consistency problem, not an established result. See the attached file AMEM-CRITICAL-REASSESSMENT.pdf for the full statement.

Version 3 (corrected) of the AMEM preprint. Primary changes from v2:

• Arithmetic correction: Baker 2025 vs Van Dyck 1987 bound ratio is ~2640 (~3.4 orders of magnitude), not 'six orders of magnitude' as stated in v2 (lines 368 and conclusions).
• Constraint table updated: Bennett et al. 2006 muon g-2 row added (|ε_s · ε_a| < 4.6×10⁻⁷, 10–100 MeV range), filling the coverage gap between the electron g-2 domain and the Baker 2025 antihydrogen bound.

The model proposes a hidden U(1)_D gauge boson coupling to matter and antimatter with different strengths, introducing C-violation in the dark sector while preserving CPT symmetry. Two experimental constraints are derived: |ε_s·ε_a| < 7×10⁻¹³ (Baker 2025, ALPHA 1S–2S, m_A' ≲ 4 keV) and |ε_s·ε_a| < 1.85×10⁻⁹ (Van Dyck 1987, e⁻/e⁺ g-2, m_A' ≲ 511 keV). The J-PET null positronium result is the expected AMEM prediction.

Companion documents: Layperson Guide (no equations, ~20 min read) and Mathematics Guide (9-chapter self-contained derivations with exercises).