Aperiodic Multiplying Media
A multiplying medium is any medium in which some quantity is generated as well as transported — neutrons in an assembly, photons in a gain stack, and, most concretely for Book 6, living cells in a growth medium. This chapter states a spectral-theoretic framework for the case where the medium is aperiodic — built by substitution rather than by periodic repetition — and shows that one number governs both its geometry and whether it grows. That number is the Perron–Frobenius eigenvalue of the medium's inflation matrix: φ for a two-symbol (Fibonacci) medium, η for a three-symbol (tribonacci) medium, climbing to τ = 2 up the recurrence ladder. It is the direct industrial descendant of the DNLS study (Chapter η · Tribonacci as Critical Constant, Book 3), which this chapter continues into the nutrition domain.
1 · The medium and its multiplication factor
An aperiodic medium is generated by a primitive substitution ρ on a finite alphabet of components A = {a₁,…,a_d}. Each component carries a transport datum (diffusion, hopping, refractive index) and a generation datum (a local growth, gain, or fission rate). The abelianised substitution is the matrix M with Mij = (number of ai in ρ(aj)). Primitivity gives, by Perron–Frobenius, a simple largest eigenvalue λPF > 1 with a strictly positive eigenvector. Under one inflation the amount of medium multiplies by λPF: it is the medium's geometric multiplication factor, present before any physics is switched on.
det M = ±1, so the inflation lives in SL(d, ℤ): the medium multiplies its component count by λPF while preserving an integer volume — the first sign that a conserved criticality rides along with the growth.
2 · The passive medium — spectrum of the transport operator
Switch on transport only. On the aperiodic component sequence the transport (Jacobi / diffusion) operator is
For primitive substitutions of this class the following are rigorous (Sütő; Bellissard; Bovier–Ghez; Damanik–Lenz), and they are the fingerprints of aperiodic order: the spectrum σ(H) is a zero-measure Cantor set (a hierarchy of gaps, no bands); it is purely singular continuous (modes neither extended as in a crystal nor exponentially localised as in disorder, but critical — algebraically localised); the integrated density of states in each gap is labelled by the frequency module ℤ + λPF−1ℤ (Bellissard gap-labelling); and the critical modes are multifractal, with inverse participation ratio scaling as IPR ~ L−D₂ at the natural inflation lengths L ~ λPFk. In short, the same λPF that multiplies components fixes the gap labels and the mode exponents. Geometry and spectrum are welded.
3 · The active medium — the criticality eigenvalue
Switch on generation. Split the operator into a transport/loss part L and a generation part F, and pose the criticality eigenvalue problem:
Below keff = 1 the population of quanta decays; above it, it grows; at 1 it is self-sustaining. Because F and L are built on the same substitution, keff is scale-covariant under inflation with a factor set by λPF, so its fixed point is pinned to the multiplication factor. Generation couples most strongly to the highest-intensity modes — the most localised critical states, those with the largest IPR.
The critical coupling (critical gain, critical nutrient contrast, or the nonlinear self-trapping threshold λc) increases with λPF because a larger λPF widens the spectral gaps (§2) and stiffens the mid-gap mode against the self-generated perturbation. Numerically: λc(n) ≈ 0.958 Δn + 0.107 across n = 2…5 (r = 0.989), with Δn the mid-gap gap width. A tribonacci medium (η > φ) therefore reaches self-generation at higher coupling than a Fibonacci one — a difference in regime, not magnitude.
4 · Nonlinear closure — DNLS as a self-multiplying medium
The DNLS equation is the case where the medium supplies its own generation; the nonlinear term is an intensity-dependent, self-dug potential:
Self-trapping is the criticality crossing of §3 with F = λ|ψ|²: below λc the mode spreads (subcritical); above it, the self-dug well binds it (a discrete breather). The V4 finding — Fibonacci self-traps in λ ∈ [4, 8] while tribonacci has not by λ = 10 — is exactly the one-number principle. The λ = 1.5 robustness (IPR retained >95% vs ~43%, an 8.6× ratio) is its linear-response precursor.
5 · Industrial application — nutrition and culture media
A growth medium is, literally, a multiplying medium: a microbial or cell population reproduces on a spatial nutrient landscape. This is the industrial lead case for Book 6, and it maps onto the framework exactly.
Fermentation & culture media on structured substrates
Linearise logistic (Fisher–KPP) growth about extinction on an aperiodically patterned substrate — a printed, layered, or scaffolded feedstock whose nutrient patches follow a substitution rule:
The linear operator D·Δ + r(x) is precisely the aperiodic Schrödinger operator of §2, and the logistic saturation −b·u² is the culture analogue of DNLS self-generation (see the companion Nonlinear Reaction-Diffusion Fold). Persistence vs. washout is decided by its principal eigenvalue crossing zero — the criticality eigenvalue of §3. Design consequence: a tribonacci-patterned substrate concentrates the surviving culture onto favourable patches more robustly than a Fibonacci one, and its persistence threshold (critical mean nutrient, or critical patch contrast before washout) sits higher — the medium tolerates leaner feed before the culture collapses. The inflation order n is a formulation knob for robustness against nutrient fluctuation.
Feed formulation & caloric criticality
The energy side of the same ledger is measured, industrially, by oxygen-bomb calorimetry (see the Parr Calorimeter tutor). The framework predicts that an aperiodically layered feed with inflation order n has a growth-criticality (minimum viable caloric contrast) that scales with λPF(n) — giving a spectral target to pair with the calorimetric one. Feed QA becomes a two-number specification: measured caloric density, and the inflation order that sets the robustness margin.
Bioreactor scale-up under inflation
Because the criticality eigenvalue is scale-covariant under inflation with ratio λPF (§3), an aperiodically structured bioreactor packing scales its criticality predictably from bench to plant — the same inflation that grows the reactor volume rescales its growth threshold by a known factor. This is the property periodic packings do not have, and it is the patentable core (Part IV · New IP). Draft invention disclosure & claim skeletons: EN · PT-BR. Investor/board briefing (bilingual): The Basis-Point Asset.
The physics siblings of these nutrition cases — photonic quasicrystal lasers, Bose–Einstein condensates in bichromatic vs. trichromatic optical lattices, and neutronically layered multiplying assemblies — obey the same one-number principle and are catalogued at the DNLS companion site. Book 6 keeps the industrial-nutrition quadrant.
6 · What is proved, measured, and open
| Statement | Status |
|---|---|
| η ∈ (1,2), η³ = η²+η+1, weight η⁻ᵏ strictly decreasing → 0 | Lean 4 · kernel-checked AXLE/TribonacciMeasure.lean, no sorry |
| Zero-measure Cantor spectrum · singular continuous · gap labelling · multifractality | classical Sütő, Bellissard, Damanik–Lenz |
| IPR robustness 8.6× · self-trapping threshold gap · λc(n) ≈ 0.958 Δn + 0.107 (r = 0.989) | numerical · V4 |
| Nutrient / bioreactor persistence-threshold ordering by λPF | predicted · untested |
| IPRtrib(0) > IPRfib(0) as a theorem · Lean threshold-law · SL(d,ℤ) ↔ modular invariance | open · AXLE roadmap |
The framework is falsifiable at every application. If an aperiodic multiplying medium reached self-generation at or below the threshold of its lower-order counterpart — a tribonacci culture washing out before a Fibonacci one under the same lean feed — the one-number principle would fail. The DNLS study is the first quadrant of that test; the nutrition media of Book 6 are the industrial rest. No claim is made that the fold thresholds here equal the series constants (τ = 2, ε₀ = 1/3, μmax = −2); those relations, if any, are stated only where separately proved.