In 2023, a neutrino with an extremely high energy level was recorded by the KM3NeT observatory in the Mediterranean Sea. The particle carried an energy around one hundred thousand times greater than the energies produced in particle accelerators on Earth. KM3NeT detected the event through a burst of Cherenkov light produced when the particle interacted with the water inside the instrumented volume. The recorded energy did not align with the typical range of neutrinos generated by known cosmic processes.
Neutrinos usually pass through matter without interaction. Detectors such as KM3NeT rely on photomultiplier arrays spaced across large volumes to capture the rare interactions that do occur. The event stood out because of its energy and because standard astrophysical mechanisms are not known to produce neutrinos at that scale. The arrival direction, timing, and energy were logged by the detector, providing the basic parameters used in later analysis.
IceCube, another neutrino observatory located at the South Pole, did not record a corresponding signal. IceCube monitors a larger volume of ice and has documented high energy neutrinos for more than a decade. Its data set includes several events above one petaelectronvolt, but none approach the scale of the KM3NeT detection. The absence of any signal near the energy recorded by KM3NeT has been noted in comparisons of the two observatories. The energy distributions in IceCube’s catalog do not include an event similar to the 2023 detection.
Researchers at the University of Massachusetts Amherst examined the discrepancy by evaluating theoretical models capable of producing neutrinos at energies recorded by KM3NeT. One model considered in their work involves primordial black holes. Primordial black holes are theoretical compact objects thought to have formed in regions of high density during the earliest moments of the universe. Their masses could vary widely, including values far below those associated with black holes formed from collapsing stars.
According to the theory of Hawking radiation, black holes emit thermal particles due to quantum effects. The rate of emission depends on the mass of the black hole. Large black holes have low temperatures and negligible emission. Smaller black holes have higher temperatures and can lose mass over time. As mass decreases, the temperature increases, leading to stronger emission near the final stage of evaporation. In the final moments, a small black hole may release particles with high energies.
The Amherst group evaluated whether a black hole undergoing its final evaporation could produce a neutrino with the recorded energy. A simple Schwarzschild black hole produces characteristic emission patterns that include a range of particle energies. These patterns did not match the combined data sets of KM3NeT and IceCube. The expected number of lower energy events did not appear in IceCube’s catalog. This conflict led the researchers to examine variations of the model.
Their study considered primordial black holes that carry a hidden charge associated with a dark sector U(1) symmetry. This charge interacts with a hypothetical heavy particle sometimes described as a dark electron. A black hole with this charge can remain close to an extremal mass configuration for long periods. In that state, evaporation slows. The spectrum of emitted particles changes relative to an uncharged black hole. The model predicts fewer neutrinos in the lower petaelectronvolt range and allows occasional high energy emissions.
The study reports that the KM3NeT neutrino energy falls within the range permitted by this model. It also reports that the lack of corresponding detections at IceCube does not contradict the scenario because the predicted flux of lower energy neutrinos would be suppressed. Calculations presented in the accepted paper indicate that the burst rates inferred from the KM3NeT and IceCube data can align within statistical limits when the dark charged model is applied.
The researchers also compared the model’s gamma ray output with constraints from gamma ray observatories. Standard primordial black hole evaporation models predict gamma ray emissions that would be detectable in certain energy bands. The dark charged model predicts reduced gamma ray output because the evaporation pathway is altered by the hidden charge. The authors’ analysis states that the absence of gamma ray detections at LHAASO is consistent with the model.
The authors note that a population of primordial black holes carrying the proposed charge could remain stable for long periods. The study states that these objects could exist in quantities large enough to account for the dark matter content of the universe. Dark matter, inferred from gravitational measurements, does not interact with electromagnetic radiation. A population of compact objects that emit little to no detectable radiation fits within some existing observational constraints. The study references this as a possible feature of the model rather than an observational claim.
The paper accepted by Physical Review Letters presents the model as one possible explanation for the recorded neutrino. It outlines parameter ranges in which the predicted emission rates match detector data. It provides calculations for the flux ratios at different energies, showing how suppression at lower energies and availability at high energies create a distribution similar to what has been observed.
The neutrino detected by KM3NeT provided the initial value used in the authors’ analysis. The event’s energy, direction, and interaction profile were included in the input parameters. IceCube’s non detection created an additional constraint that the model needed to match. Gamma ray observatories provided further constraints. The authors state that the combined dataset can be satisfied under the dark charged primordial black hole scenario.
No direct evidence of primordial black holes has been recorded by any observatory. Hawking radiation has not been directly observed. The model described in the paper represents a theoretical interpretation applied to a single neutrino event combined with constraints from multiple detectors. The authors identify ranges of mass, charge, and emission properties for which their model fits existing measurements. These values are then compared against broader astrophysical limits to test compatibility.
Neutrino observatories are expected to continue collecting data. KM3NeT is still expanding its instrumented volume. IceCube continues to record events at a high rate. Additional high energy neutrino events could allow further tests of the model. The authors note that a consistent pattern of neutrinos at similar energies would provide more opportunities to evaluate their framework. They also note that the absence of such events over extended periods would place limits on the model’s parameter space.
The accepted paper focuses on the properties of the hypothetical black holes and the resulting neutrino fluxes. It does not claim direct detection of a primordial black hole. It reports that the model provides a possible explanation for the features of the 2023 event. It includes mathematical treatments of emission spectra, burst rates, and detector sensitivities. These calculations form the basis of the authors’ interpretation of the KM3NeT and IceCube data.
The KM3NeT event remains the highest energy neutrino recorded to date. Its origin has not been linked to any known astrophysical source. Observatories that monitor the relevant regions of the sky have not reported signals that match the energies associated with typical high energy cosmic events. The datasets remain limited to the neutrino detection, IceCube’s non detection, and the absence of gamma rays linked to the direction of the event.
The study presents one framework that accommodates these data. Other models remain under consideration in the wider scientific community. The unusual energy of the event has made it an object of interest for researchers studying high energy astrophysics, particle physics, and dark matter theory. Further observations will determine whether additional events share similar characteristics.
Source:
Baker, M. J., Iguaz Juan, J., Symons, A., and Thamm, A.
“Explaining the PeV neutrino fluxes at KM3NeT and IceCube with quasiextremal primordial black holes.”
Physical Review Letters, accepted December 18, 2025.
DOI: https://doi.org/10.1103/r793-p7ct
