This paper reports preliminary results from an experiment that used gamma rays to count which atomic nuclei appear when uranium-235 is split by slow (thermal) neutrons. The team ran the FIPPS gamma-ray spectrometer at the Institut Laue‑Langevin and used an “active” target: uranium-235 dissolved in a liquid scintillator. That target both started the fission and signalled each fission event by producing light, which let the researchers separate prompt gamma rays (emitted immediately by the fission fragments) from delayed gamma rays (emitted later by radioactive decay).

To get absolute, independent fission yields, the researchers looked at the prompt gamma rays that identify specific, even-even fission fragments. They used two related analysis methods. One (M1) sums the gamma-ray lines that feed a fragment’s ground state. The other (M2) uses coincidences between successive gamma rays to follow likely decay cascades. The experiment recorded about 9,095 five-minute runs and used timing and amplitude cuts on the photomultiplier tube signal to sort prompt from delayed events. The team also applied corrections for detector dead time, signal summing, contamination, isomeric transitions, and spectrometer calibration.

Most measured yields agreed well with the standard evaluated database JEFF‑3.3. A striking exception was the nucleus tin-132 (132Sn). The measured yield for 132Sn was 0.107(5), which is about five to seven times lower than the JEFF‑3.3 value (0.74(19)) and the ENDF/B‑VIII value (0.59(2)). Using the FIFRELIN fission-fragment de-excitation simulation, the authors interpret this deficit as evidence that 132Sn is often produced directly in its ground state at scission (the moment the nucleus splits) or after neutrons are emitted. In one FIFRELIN simulation variant, about 66% of 132Sn appeared directly in the ground state at scission and another 14% directly after neutron evaporation. That “ground-state production” would not produce the prompt gamma rays the experiment counts, so the observed prompt-gamma yield is lower even if the real production is higher.