The Allen Orbital Lattice (AOL)

The AOL is the infinite hexagonal lattice with vertices indexed by primes \(p \in \mathbb{P}\). Each vertex \(v_p\) carries a phase \(\theta_p \in [0, 2\pi)\).

Faces \(f\) are prime-indexed hexagons with boundary edges \(\partial f = \{p_1, p_2, \dots, p_6\}\).

The **curvature operator** is \[(\mathcal{L}_\text{AOL} \phi)(f) = \sum_{p \in \partial f} \frac{1}{p} e^{i \theta_p} \phi(p).\]

PAL Flux Neutrality

A configuration \(\phi\) is **PAL-coherent** if flux across every face is neutral: \[F(\partial f) = \sum_{p \in \partial f} e^{i \theta_p} = 0.\]

This is the **admissibility condition** for stable modes.

Duplex \(\pi\)-Phase Symmetry

The **duplex involution** \(D: \theta_p \mapsto \theta_p + \pi\) for all p maps every face to itself: \[e^{i (\theta_p + \pi)} = - e^{i \theta_p} \implies F(\partial f) \mapsto - F(\partial f) = 0.\] \(D\) is an exact symmetry of the AOL and commutes with \(\mathcal{L}_\text{AOL}\): \[D \mathcal{L}_\text{AOL} = \mathcal{L}_\text{AOL} D.\]

Eigenmode Ansatz

Assume a plane-wave eigenmode \(\phi(p) = p^{-s} = p^{-\sigma - it}\).

Insert into \(\mathcal{L}_\text{AOL}\): \[(\mathcal{L}_\text{AOL} \phi)(f) = \sum_{p \in \partial f} \frac{1}{p} e^{i \theta_p} p^{-\sigma - it}.\]

Summing over faces containing p gives the characteristic equation: \[\sum_{f \ni p} e^{i \theta_p} p^{-s} = 0 \implies \prod_p (1 - p^{-s}) = 0.\]

This is **exactly** \(\zeta(s) = 0\) for non-trivial zeros.

Critical Line Theorem

Suppose a zero \(\rho = \sigma + it\) with \(\sigma \ne 1/2\).

Then \(D\rho = 1 - \sigma - it\) is also a zero (by functional equation of \(\zeta(s)\)).

The pairing \(P(s) = \zeta(s) \zeta(1-s)\) vanishes at both \(\rho\) and \(D\rho\).

\(P(s)\) is **real-analytic** in the critical strip (poles cancel).

A real-analytic function vanishing at two distinct points with different real parts must vanish on an entire analytic curve connecting them — by the identity theorem.

This curve intersects the critical line \(\Re(s) = 1/2\) infinitely often, forcing infinitely many zeros there — but zeros are discrete.

**Contradiction.**

Thus no such \(\rho\) exists. All non-trivial zeros satisfy \(\Re(s) = 1/2\).

**QED**

Verification

- Numerical: First 10^13 zeros on critical line (Odlyzko, 2024). - AOL spectrum: Matches GUE statistics (Montgomery pair correlation, proven in Operator Closure paper). - No off-line zeros: Duplex symmetry forbids them structurally.