The paper studies a possible dark-matter explanation for one unusual event reported by the LUX‑ZEPLIN (LZ) experiment. LZ saw a single nuclear recoil with reconstructed energy about 248 keV. That high energy is hard to explain with ordinary elastic dark-matter scattering, so the authors ask whether an inelastic dark-matter model can reproduce the event while also matching how much dark matter we see in the universe.

The model they use is a pseudo-Dirac fermion. In plain language, the dark particle comes in two almost identical states separated by a tiny mass gap. When a dark particle hits a nucleus it can be kicked into the slightly heavier state. That “endothermic” process uses some incoming kinetic energy to make the heavier partner. Because of this energy cost, low-energy recoils are suppressed and the scattering tends to produce fewer low-energy events and relatively more high-energy recoils — exactly the feature that makes the 248 keV event interesting.

The cosmology behind the model is also important. Instead of the usual thermal freeze-out idea, the authors consider freeze-in production with a low reheating temperature after the early universe’s hot phase. In freeze-in the dark sector never reaches thermal equilibrium and its abundance builds up from rare production processes. If the reheating temperature is low, dark-matter production is Boltzmann suppressed and the correct relic abundance can be achieved with stronger couplings than in the usual high-temperature freeze-in case. That extra freedom means the reheating temperature becomes a new parameter. It breaks the one-to-one link between dark-matter mass and detection rate that thermal freeze-out enforces. As a result, the LZ candidate can be matched across a continuous region of model parameters rather than at a single “thermal” target.