18.1 Two Operators, One Cage
A zeolite is a crystalline aluminosilicate threaded with molecular-scale channels — pores measured in angstroms, the same unit used for atoms themselves. ZSM-5, the workhorse of the petrochemical industry, has the MFI framework: two intersecting channel systems built from 10-membered oxygen rings, with apertures of roughly 5.3 × 5.6 Å (straight channels) and 5.1 × 5.5 Å (sinusoidal channels). Anything that wants to pass through has to fit.
Inside those channels sit the catalytic sites — Brønsted acid centers, or confined metal clusters such as nickel or ruthenium. This gives every zeolite-confined reaction two operators acting on the same molecule, in an order that matters:
Kψ = θ(η* − d(ψ)) · ψ — a Heaviside projection. A species ψ with kinetic diameter d(ψ) passes only if d(ψ) < η*, the pore aperture (η* ≈ 5.5 Å for MFI). This is shape-selective catalysis in its purest form: the cage is a literal molecular sieve.
Fψ = ψ + λ·R(ψ) — the confined catalytic transformation itself: bond-breaking, bond-forming, a Whitney-fold of the reaction coordinate at the metal cluster or acid site. R(ψ) can change what the molecule is — and therefore change d(ψ), the very quantity K tests.
| Species | Approx. kinetic diameter | Fits MFI pore (η* ≈ 5.5 Å)? |
|---|---|---|
| H₂ | 2.9 Å | always |
| CO₂ | 3.3 Å | always |
| CH₄ | 3.8 Å | always |
| Light olefins (C₂–C₃) | 3.9–4.5 Å | usually |
| p-xylene | ≈5.8 Å | marginal — the classic ZSM-5 selectivity case |
| o-/m-xylene | ≈6.8 Å | excluded |
Diameters are approximate Lennard-Jones kinetic diameters from the standard zeolite-adsorption literature; the point is the ordering, not the third decimal place.
18.2 Theorem 18.1 — The Cage Does Not Commute
TOTOGT/io → ZeoliteCommutation.lean.
Theorem 18.1 below does not depend on that formula and is unaffected. Its gate is not fixed: K = θ(η* − d(ψ)) is threshold-state-dependent, and F changes d(ψ). A gate whose argument the fold rewrites is a different object from a static mask, and it genuinely fails to commute. The mechanism was always the state-dependence; the δ was never doing the work.
Zeolite confinement is order-dependent for a concrete and checkable reason. The pore admits a species by its kinetic diameter d(ψ); the confined reaction changes that diameter. Whether the gate is consulted before or after the reaction therefore decides which species exist to be gated — and that is exactly what the two orderings below produce. Order-dependence here is carried by the operator that alters the gated property, never by a static aperture acting alone.
18.3 Methane: The Fixed-Point Molecule
One molecule sits at d(ψ) ≈ 3.8 Å — comfortably below η* ≈ 5.5 Å under any ordering. Tetrahedral, nonpolar, the smallest possible carbon-hydrogen unit: methane (CH₄) is admitted by K whether F acts before or after it. It is, in the language of Chapter 16, a fixed point of the commutator — the one product the cage never argues about.
That is not a coincidence the Mini-Beast can ignore. G = U∘F∘K∘C was shown in Theorem 16.1 to be a fixed point of the renormalization map — the same structure at every scale. Methane is the molecular-scale echo of that statement: the smallest stable hydrocarbon is also the one species for which [K,F] effectively vanishes. Where the operator algebra disagrees about everything else, it agrees about methane.
EN. Theorem 18.1 is the interesting case precisely because its gate is not fixed. The finite refutations kernel-checked in ZeoliteCommutation.lean treat a static 0/1 gate; here K = θ(η* − d(ψ)) has a threshold argument that F itself rewrites. That is a genuinely harder object to machine-check, and nobody has yet.
Open problem D-ZEO′ (open · full credit for a proof or a disproof). Model a discrete species set with an integer "diameter" tag on each site, a gate that keeps sites whose tag is below η*, and a reaction F that changes the tag (reactant tag → product tag). Prove in Lean, on a `Fin N` state, that [K, F] ≠ 0 exactly when some species is carried across the threshold by F (reactant fits, product does not, or vice-versa) — and that it vanishes when F preserves the side of η*. This is the state-dependent-gate analogue of gate_fold_not_commute: the first machine-checked commutation result where the gate reads a property the fold mutates.
On-ramp: the static-gate finite witness (riboswitch Open Problem D1-T1′) is the warm-up; the new content is making the gate's argument depend on the state F returns. Style: pattern-match on Fin N, never ![…]; verify with #print axioms.
18.4 The Sabatier Bridge — C and F Together
The reaction that makes methane from the simplest possible feedstock has a name: the Sabatier reaction, known since 1897.
C, the constraint operator, is the gathering step: concentrating a dilute CO₂ atmosphere into a pressurized reactor feed, with H₂ supplied by water electrolysis. On Mars — 96% CO₂ at roughly 0.6 kPa surface pressure, with water ice in the regolith — this is exactly the resource picture. A 2011 NASA prototype Sabatier reactor ran continuously for five days on simulated Martian atmosphere, producing methane at near-100% CO₂ conversion at a rate of about 1 kg/day. Zeolite-confined nickel catalysts now reported in the literature push CH₄ selectivity toward 100% at moderate temperatures (≈280 °C) — the K-operator doing real chemical work, not just metaphorical work.
18.5 U — Unfolding into Thrust
The unfold operator U is where the chain leaves the laboratory and becomes propulsion. CH₄ + 2 O₂ → CO₂ + 2 H₂O releases its stored bond energy in combustion — and "methalox" (liquid methane + liquid oxygen) is the propellant pairing behind SpaceX's Raptor engines, which power both stages of Starship. The choice was explicit and ISRU-driven from the start: methane is the one major rocket fuel that can be manufactured in situ on Mars by running the Sabatier reaction in reverse-engineered form — CO₂ from the air, H₂ from electrolyzed water-ice, out comes CH₄ and O₂ for the return trip.
Trace the full chain on a single tank of Martian propellant: C concentrates CO₂ and water ice into reactor feed; K confines the reaction to a zeolite cage sized to admit only small molecules; F folds CO₂ + H₂ into CH₄ + H₂O at the confined catalytic site; U unfolds the stored chemical energy of CH₄ + O₂ into thrust. G = U∘F∘K∘C, instantiated in steel and aluminosilicate, is a return ticket from another planet.
18.6 Interactive: The Molecular Sieve
The bar chart plots kinetic diameter against the MFI pore aperture η* ≈ 5.5 Å. Bars below the threshold line are admitted by K under either ordering; bars that cross it are exactly where [K,F] ≠ 0 becomes chemically consequential — where the order of confinement and reaction decides the product.
⊞ Kinetic Diameter vs. Pore Aperture (MFI, η* ≈ 5.5 Å)
18.7 The g-Series of Way-Stations
Chapter index follows the g-series of recurrence cycles. Reading it as a roadmap for in-situ propellant production gives the chain a direction — and a deliberate stopping point.
| Regime | ISRU analogue | Status |
|---|---|---|
| g⁰ — Quiescent | CO₂/H₂O ice reservoir, unreacted (Mars regolith, cometary ice) | observed |
| g² — Nascent oscillation | First Sabatier plant cycling CO₂+4H₂⇌CH₄+2H₂O on a planetary surface | prototyped (NASA, 2011) |
| g⁶ — Stable micro-cycle | A small network of depots completing supply loops (Mars + orbital staging) | engineering target |
| g³³ — Stability threshold | An outer-system relay chain (Europa, Enceladus, Titan — each carrying CO₂/H₂O/CH₄ reservoirs) where one G-pass is routine | not yet attempted |
| g⁶⁴ — Circuit saturation | "Intergalactic expanse" — a propellant lattice spanning beyond the solar system | Axiom 9 — honest incompleteness |
That last row is written deliberately, not as a promise. Axiom 9 of the dm³ framework — Honest Incompleteness — says the recurrence ladder π → φ → μ → η → Δ → Σ → Ω → π converges toward the embodiment threshold τ = 2, and that this is the only known scale-invariant template if a propellant lattice of that scale is ever built. The chapter does not claim g⁶⁴ exists. It claims that the same non-commuting K and F that decide whether a single zeolite cage makes methane or coke also decide, recursively, whether each way-station along such a chain produces a usable G-pass or a dead reactor. The algebra is scale-invariant even where the hardware is not yet built.
18.8 Open Problem for the Camarada
- Sabatier reaction — overview and thermochemistry
- NASA NTRS — Mars Atmospheric Conversion to Methane and Water (Sabatier reactor, Ru/Al₂O₃)
- National Science Review — Cavity-controlled methanol conversion over zeolite catalysts
- PMC — Engineering of Transition Metal Catalysts Confined in Zeolites
- SpaceX Raptor — methalox engine, ISRU design rationale
- Frontiers — Ni-doped ETS-10 zeolite for CO₂ methanation