Historical research guide · 0481 · 2025
Anomalous Heat Reaction from Hydrogen and Metals
Tadahiko Mizuno · Jed Rothwell
In one minute
Tadahiko Mizuno has spent decades in Sapporo measuring heat from hydrogen in metals, and this paper reports something he found by accident: the plain steel reactor he was using as a control was itself producing heat. Chasing that, he and Jed Rothwell tested five stainless-steel reactors in an air-flow calorimeter — an insulated box with a measured stream of cooling air through it, so you can compare the heat carried out against the electricity put in. Four of the five gave out more than they took in, once the heat escaping through the chamber walls is added back: 115 watts out for 100 watts in on one reactor, 689 for 601 on the same one at high power, and as much as 285 watts of excess from another. The one reactor that never received the surface treatment produced none. The heat climbs exponentially with temperature, along a straight Arrhenius line whose slope matches the energy it costs hydrogen to move through nickel.
Connection
Chapter 12 holds that a hydrogen-loaded metal lattice does real work at ordinary temperatures, and this is that claim reduced to its cheapest form: no palladium, no heavy water, no electrolysis — a stainless-steel tube whose inner surface has been buffed to carry defects, hydrogen or ordinary air, and a heater. The paper’s most useful content for anyone building one is the preparation recipe and the Arrhenius slope, which points at hydrogen diffusion in the metal as the rate-setting step, and therefore at what to change next.
From the source
A stainless-steel reactor whose surface has been suitably treated produces thermal energy exceeding the electrical energy put into it. Reactor R40, a SUS304 tube 400 mm long holding a 27 gram nickel mesh under hydrogen at 1000 pascals, gave 115.1 watts of corrected output for 100.0 watts of input, and 688.7 watts for 601.0 watts of input, an output-to-input ratio of 1.15.Section 7, Test Reactor R40; Section 8, Test Results; Figures 10 and 11
Four of the five reactors in the study produced excess heat, with output-to-input ratios from 1.114 to 1.379 and excess power as high as 285 watts for the cross-shaped R36; the exception, R35, is the one reactor with no discharge electrode and therefore no activation treatment, and it produced none.Section 8, Table 2; Section 10, Conclusions, items 1 to 4
The measurement is air-flow calorimetry with an explicit heat-loss correction, because not all the heat leaves through the airstream: at a 603 watt calibration about 400 watts are carried by the air and about 200 watts escape through the chamber walls, and the loss is fitted as an exponential function of the outlet-minus-inlet air temperature. The claim also survives in raw temperatures — reactor R36 at 504 watts input raised the outlet air 31.7 degrees Celsius above inlet, against 27.5 degrees for a 501 watt calibration heater, a difference of 4.1 degrees.Sections 5 and 6, Calorimetry and Calibration Method; Table 1; Equations 11 and 12; Figures 7, 15 and 16
Excess heat rises exponentially with reactor temperature and falls on a straight line against the reciprocal of absolute temperature, giving activation energies between 0.23 and 0.29 electronvolts per atom across four reactor shapes — very close to the 0.298 electronvolts per atom activation energy for hydrogen diffusion in nickel, which the authors read as evidence that the amount of dissolved hydrogen is the important factor.Section 8; Figures 12, 13 and 17; the comparison with the nickel diffusion activation energy
The preparation is the experiment. The metal surface must be cleaned and mechanically stressed to create defects a few micrometres deep and a few nanometres across — buffing works, electropolishing and emery polishing do not because they remove the worked layer — after which the reaction runs in hydrogen, in deuterium above 350 degrees Celsius, or in open air; and electrical discharge treatment in low vacuum produces the extremely clean, hydrogen-active surface that intensifies it. Without the defects there is no reaction.Section 9, Material Preparation, items 1 to 10
Four independent confirmations are named: Sanshu Kogyo built its own reactor and measurement system and confirmed the excess heat; Dynax Industries confirmed it in a thermostatic chamber they built themselves; Jean-Paul Biberian at the University of Marseille performed a preliminary confirmation on a reactor supplied by Mizuno; and Murata Manufacturing, using a transpiration heat measurement system at a new research site, confirmed 200 watts, more than three times input, maintained for several weeks across seven reactors supplied by Mizuno. Those are the results to watch, because scale and duration are what turn this from an effect into a power source.Acknowledgements
The way in
https://doi.org/10.70923/001c.134027How to cite it
Tadahiko Mizuno, Jed Rothwell (2025) Anomalous Heat Reaction from Hydrogen and Metals doi:10.70923/001c.134027