The lattice ladder

Why the library suggests it

The NASA Glenn team loaded erbium and titanium with deuterium, hit the samples with a bremsstrahlung beam of no more than 2.9 million electronvolts, and measured 2.45 MeV neutrons — the fingerprint of deuterium-deuterium fusion — repeatably, from fuel that never moves (Novel nuclear reactions in bremsstrahlung-irradiated deuterated metals, 2020). The programme around that measurement, including the four-panellist review whose job was to attack it, is set out in Lattice Confinement Fusion Overview (2020) and in plain language for engineers in A New Way of Triggering Nuclear Fusion (2022), which gives the loading figure that makes it work: about ten to the twenty-third deuterons per cubic centimetre, nine orders of magnitude denser than a magnetically confined plasma. The size of the screening effect is not a model, it is measured on an accelerator: in titanium the screening energy is about 27 electronvolts, in palladium about 310, in palladium oxide about 600 — and at a bombarding energy of 2.5 kiloelectronvolts, palladium oxide fuses about a hundred times more often than the bare-nucleus rate (Screening energy for low energy nuclear reactions in condensed matter, 2020). The NASA and Navy account of the mechanism, matched against a measured shift in beryllium-7's half-life when caged inside a buckyball, is Electron Screened and Enhanced Nuclear Reactions (2022).

The experiment or build

The next rung is not a bigger machine, it is a materials series. Kasagi's screening energies were measured on a handful of hosts. Nobody has run the full sweep: load twenty different hosts and host chemistries — metals, oxides, alloys, nanostructured films — to a measured stoichiometry, fire the same low-energy deuteron beam at each, and publish the screening energy against the material. The settling measurement is screening energy in electronvolts, per host, with the loading ratio measured independently, plotted so that a materials engineer can read off which solid to build a reactor from. That plot does not exist and it is the single most useful missing document in the field.