One fission run, fifteen nuclei: physicists map a stubborn gamma-ray excess

At 21:40 UTC on 16 July 2026, the American Physical Society's Physics magazine published a study reporting that physicists have, in a single experimental run, mapped the high-energy gamma-ray emissions of more than a dozen unstable heavy nuclei produced by induced fission. The measurement, carried out at the ALTO facility of the Institut de Physique Nucléaire d'Orsay in France, is the broadest simultaneous survey of so-called "pygmy" and higher-energy dipole strength in short-lived fission fragments ever attempted in one campaign, and it returns a long-standing puzzle to the centre of nuclear-structure research.

The puzzle is roughly forty years old. Nuclear physicists have known since the 1980s that their models under-predict the number of low-energy photons that unstable, neutron-rich nuclei emit. The extra emission is small in absolute terms, but it is large relative to what theory expects, and it sits at exactly the energies that govern how exotic nuclei absorb or shed energy inside stars and reactors. Without a quantitative handle on it, models of element formation in neutron-star mergers and of heat generation in next-generation reactors carry a stubborn residual uncertainty. The Orsay result does not resolve the discrepancy, but it puts a much sharper ruler next to it: fifteen nuclei, measured under identical conditions in one beam time, where previous surveys cobbled together data from a decade of separate runs.

What the experiment actually did

The team fired a beam of fast neutrons into a thick actinide target, inducing fission and producing a spray of neutron-rich fragments across the periodic table. They then caught those fragments, separated them by mass, and implanted them at the centre of a compact array of high-purity germanium detectors tuned to register gamma rays in the energy range where the excess has historically been reported. By repeating the cycle for many hours, the collaboration accumulated enough statistics on each isotope to extract a clean photon-strength function, the curve that tells physicists how readily a nucleus radiates gamma rays at a given energy.

The published result spans isotopes from the neighbourhood of tin-132 out past barium and into the lanthanide region, with masses covering both the doubly-magic tin endpoint and more deformed mid-shell nuclei where the original discrepancies were first noted. The same detector geometry, the same neutron flux, and the same analysis pipeline were applied across the board. That uniformity is the contribution: it converts what was a patchwork of single-isotope measurements, often taken with different detectors at different facilities, into a coherent internal comparison.

What the data are saying, and what they are not

Read across all fifteen nuclei, the measured photon-strength curves sit above the predictions of standard nuclear-structure models in the same energy window where the historical excess lives. The agreement with theory is not bad in absolute terms; the curves have the right shape. The disagreement is quantitative, and it is systematic: the models consistently under-predict the measured yield, sometimes by a factor of two or more at the energies of interest.

There are at least two ways to read that. The dominant reading, the one the Orsay team effectively endorses, is that the excess is real and physical. It would mean that the wave functions of these very neutron-rich nuclei contain collective modes, possibly a soft dipole oscillation of the neutron skin against the nuclear core, that the standard shell-model-inspired models under-describe. The alternative reading is that the discrepancy is, at least partly, an artefact: incomplete treatment of level density in the statistical part of the calculation, or an underestimate of feeding from higher-lying states populated in the fission process. The new data sharpen the comparison enough that both camps can now be tested against the same yardstick. They do not yet adjudicate between them.

Why one experiment across many nuclei matters

Until now, the field's working assumption was that the excess scales with the neutron-to-proton ratio of the nucleus. The Orsay data complicate that assumption. The measured enhancement is not a smooth function of neutron excess; isotopes with similar neutron skins show different degrees of extra strength, and some nuclei with relatively modest skin thickness display an excess comparable to nuclei with much thicker ones. That argues against a single, geometry-driven mechanism and in favour of a more nuanced picture in which shell structure and pairing correlations do substantial work.

For applications, the immediate beneficiary is nuclear astrophysics. The same neutron-rich nuclei are produced in the r-process, the rapid neutron-capture chain responsible for roughly half the heavy elements beyond iron, including the gold in wedding rings and the iodine in human thyroid glands. The rates at which those nuclei capture further neutrons, or undergo photodisintegration, depend sensitively on the photon-strength functions the Orsay experiment has now re-measured. Cleaner curves feed directly into cleaner predictions of how much of each element a given astrophysical site should produce.

Reactor physics is a quieter but real beneficiary. Next-generation fast reactors and accelerator-driven systems burn fuel cycles that pass through several of the isotopes in the new dataset. Their heat-generation models include a contribution from the same low-energy gamma emission the new measurement has re-characterised, and small changes in that term propagate into safety margins and fuel-loading decisions.

Where the field goes from here

The campaign at ALTO is not finished. The collaboration has flagged follow-up measurements with a larger germanium array, longer beam times, and heavier actinide targets that should extend the dataset toward even more neutron-rich isotopes and toward the heavier actinides where the statistical-model assumptions are most strained. The wider community, meanwhile, is already folding the new photon-strength functions into its r-process network calculations; revised abundance predictions should appear in preprints within the next several months, and the first peer-reviewed updates are likely before the end of 2026.

What remains genuinely uncertain is how the new curves will land inside the models. The standard approach treats the low-energy photon strength as a smooth extrapolation from the giant dipole resonance that dominates at higher energies. The Orsay data suggest the extrapolation needs more structure than current parametrisations allow, but they do not, on their own, specify what that structure should be. That is the next paper's job, and the field is now better equipped to write it.

This desk covered the result as a measurement advance rather than a model refutation: the experiment sharpens the comparison between data and theory across a sample of fifteen nuclei in a single campaign, but the interpretation of the residual excess remains contested. Wire coverage from Physical Review C and APS Physics has framed it similarly.