Historical research guide · 0863 · 2025
Photon Radiation Calorimetry for Anomalous Heat Generation in NiCu Multilayer Thin Film During Hydrogen Gas Desorption
J. Kasagi · T. Itoh · Y. Iwamura · Y. Shibasaki · T. Takahashi · S. Yamauchi
In one minute
Heat is hard to measure honestly at a thousand kelvin, and this field has been caught out before by emissivity errors. Jirohta Kasagi and colleagues at Tohoku University, working with Clean Planet, answer that by counting photons instead of trusting a thermometer. Their samples are thin nickel-copper multilayer films — layers a few tens of nanometres thick — loaded with hydrogen gas and then heated to about 1100 kelvin in vacuum while the hydrogen comes back out. Three detectors watch the emitted light from 0.22 to 4.13 electron-volts, and the spectra fit a grey-body curve cleanly enough that surface temperature and emissivity can be read off separately. That separation matters, because hydrogen changes the emissivity and that change alone can imitate a temperature rise. Fed into a heat-flow model, the nickel-copper films run 4 to 6 watts hotter than their own hydrogen-free calibration. Over eighty hours that comes to 460 kilojoules, at least 410 keV per hydrogen atom, which the authors say is not a chemical reaction at all.
Connection
Chapter 12 is the case that a metal lattice loaded with hydrogen can host energy release at nuclear scale, and this is the measurement discipline that case needs: a calorimeter that can tell a real power increase from an emissivity change, applied to a sample producing whole watts. For chapter 6 the number to hold on to is the one per atom — 410 keV is roughly five orders of magnitude above anything the electron shells can supply.
From the source
Photon radiation calorimetry has been developed for these measurements: three types of photon detector cover wavelengths from 0.3 to 5.5 micrometres, that is photon energies from 0.22 to 4.13 electron-volts, reading radiant power from both faces of the sample. The chamber is held at ten to the minus four pascal or less, so the mean free path of the residual gas is over ten metres and every heat path except thermal radiation and a slight conduction can be neglected.Abstract; Section 1, Introduction; Section 2.1, Experimental Setup
Four samples are compared: a pure nickel foil, a Ni5Cu1 film of six copper-nickel bilayers sputtered on a nickel substrate, a Ni1Cu3 film of the same bilayer count with the thicknesses reversed, and a single 140 nanometre copper layer on nickel. Each is baked in vacuum for about three days near 1150 kelvin, then loaded with hydrogen gas at about 520 kelvin and measured while the gas desorbs under heating.Sections 2.2 and 2.3, Samples and Experimental Procedure
The measured spectra are well described by grey-body radiation, so emissivity and surface temperature can be fitted separately. This is the crux of the method: introducing hydrogen lowers the emissivity, mainly by reducing the surface oxide layer, and a ten per cent fall in emissivity raises the surface temperature by about two per cent even when no excess power at all is generated — which is exactly how a thermometry-only calorimeter is misled.Section 3.1, Eq. (1) and Table 1; Section 3.2, Evaluation of Excess Heat
Every sample radiates more strongly during hydrogen desorption, and the nickel-copper composite films radiate very much more: with the emissivity correction carried through the heat-flow model, they produce about 4 to 5 watts of excess heat power, up to 17 per cent of the heater input, while the pure nickel sample with hydrogen is almost indistinguishable from the same sample without it.Section 3.1; Section 3.3, Quantitative Comparison of Excess Power, Fig. 5
In an eighty-hour continuous run on the Ni1Cu3 sample the total energy generated is 460 plus or minus 120 kilojoules, from six millionths of a mole of hydrogen molecules absorbed in the sample: at least 410 plus or minus 108 kilo-electron-volts per hydrogen atom. That is far beyond any energy exchange available in electronic levels, so it is definitely not a chemical reaction; it produces energy at the level of nuclear reactions. The true figure per atom should be larger still, since most of the absorbed hydrogen leaves by thermal diffusion without taking part.Section 3.4, Generated Energy, Fig. 6; Section 4, Conclusion
The role of the initial layer structure is still open: no trace of the layered structure survives the three-day bake at about 1150 kelvin, so what the nickel-copper pairing actually does is not yet settled, and the relationship between the magnitude of the generated heat power and both the nickel to copper ratio and the film structure remains to be addressed in future work.Section 3.3, final paragraph
The way in
https://doi.org/10.70923/001c.134004How to cite it
J. Kasagi, T. Itoh, Y. Iwamura, Y. Shibasaki, T. Takahashi, S. Yamauchi (2025) Photon Radiation Calorimetry for Anomalous Heat Generation in NiCu Multilayer Thin Film During Hydrogen Gas Desorption doi:10.70923/001c.134004