On February 13, 2023, something slammed into the Mediterranean Sea off the coast of Sicily with roughly 100,000 times more energy than anything produced by the Large Hadron Collider. No explosion followed. No tsunami formed. The particle—a neutrino—passed through the planet as easily as light through a window, but not before triggering sensors in the KM3NeT detector array. At 220 petaelectronvolts, it was 35 times more energetic than any neutrino ever recorded.
The problem? Nothing in the known universe should be producing neutrinos like this.
The Ghost Particle Problem
Neutrinos are everywhere. About 100 trillion pass through your body every second, most originating from the sun. They're nearly massless, electrically neutral, and interact so weakly with matter that they've earned the nickname "ghost particles." Detecting even one requires building massive instruments and waiting patiently.
IceCube, the premiere neutrino observatory buried in Antarctic ice, has been running for over a decade. It's found hundreds of cosmic neutrinos and five above 1 petaelectronvolt. But nothing approaching 100 PeV, let alone 220. The discrepancy between what KM3NeT saw and what IceCube should have detected created what physicists call a "3.5 sigma tension"—statistical speak for "something doesn't add up."
When researchers catalogued every known astrophysical source capable of generating ultra-high-energy neutrinos—supernovae, blazars, gamma-ray bursts—none fit the energy profile. The particle seemed to come from nowhere.
Black Holes That Never Grew Up
Enter an idea from 1966, when Soviet physicists Yakov Zel'dovich and Igor Novikov proposed that black holes might have formed in the universe's first split second, before stars or even atoms existed. Stephen Hawking refined the concept in 1971, showing that these primordial black holes could be vastly smaller than their stellar cousins—some as tiny as an atomic nucleus.
Unlike stellar black holes, which form from collapsing stars and measure at least a few solar masses, primordial black holes could span an enormous range. The sweet spot for ones that might still exist today runs from about 10^17 grams (an average asteroid) to 10^23 grams (a moon). Anything lighter would have already evaporated through Hawking radiation.
That evaporation process follows a counterintuitive rule: the smaller the black hole, the hotter it burns. As Dr. Andrea Thamm of the University of Massachusetts Amherst explains, "The lighter a black hole is, the hotter it should be and the more particles it will emit." This creates a runaway effect. As the black hole radiates energy, it loses mass, heats up, radiates faster, and eventually explodes in a final burst containing all 17 fundamental particles in the Standard Model.
A black hole massing 100 trillion grams has a lifetime roughly equal to the age of the universe. We should be seeing some explode right now.
The Dark Charge Solution
But here's where the math gets interesting. Thamm and her colleagues—Michael J. Baker and Joaquim Iguaz Juan—weren't satisfied with standard primordial black hole explosions. The energy spectrum was wrong. These explosions should produce plenty of neutrinos at 1 PeV, exactly where IceCube excels at detection. Yet IceCube saw nothing while KM3NeT detected an event 100 times more energetic.
Their solution involves what they call "quasi-extremal primordial black holes" with a "dark charge." This isn't electric charge as we know it, but something analogous operating in a hidden sector of physics—a mirror version of electromagnetism involving hypothetical particles called dark electrons.
When a primordial black hole acquires enough dark charge, something remarkable happens. The electrical repulsion between like charges begins to balance gravity's inward crush. The black hole enters a cosmic coma, nearly ceasing its evaporation and becoming "cosmologically long-lived." It sits dormant, possibly for billions of years.
Eventually, the intense dark electric field at the black hole's surface tears space-time itself apart through what's called the dark Schwinger effect, creating pairs of dark electrons from pure energy. This triggers a rapid discharge and explosion—but one with a very different particle signature than a standard primordial black hole death.
The beauty of this model is specificity. The quasi-extremal explosion suppresses neutrino emission precisely at 1 PeV while allowing it at 100 PeV. It explains both what KM3NeT saw and what IceCube didn't.
Exploding Black Holes as Dark Matter Detectors
If this explanation holds, the implications extend far beyond one anomalous particle. These quasi-extremal primordial black holes could constitute all of the universe's dark matter—the 85% of matter we know exists but have never directly observed. A significant population of them would be consistent with other astrophysical observations, and their explosions would provide the first direct evidence of dark sector particles.
David Kaiser at MIT, who helped propose the primordial black hole explanation (though he admits he was "half-joking" at first), points out that we'd be using one cosmic mystery to solve another. The same objects explaining the ultra-high-energy neutrino could explain why galaxies rotate faster than visible matter alone predicts.
When the Next One Arrives
The team published their findings in Physical Review Letters in December 2025. Now the wait begins. If quasi-extremal primordial black holes are exploding throughout the universe, KM3NeT should detect more ultra-high-energy neutrinos with similar characteristics. IceCube should continue seeing nothing at lower energies. Other detectors coming online in the next decade will either confirm the pattern or force physicists back to their whiteboards.
Mansi Kasliwal at Caltech, who previously linked a high-energy neutrino to a tidal disruption event in 2019, sees this as the maturing of multimessenger astrophysics. "First, it was a blazar and now it is a tidal disruption event," she noted after that discovery. Perhaps next it will be the explosive death of a black hole older than stars, carrying evidence of particles that have hidden in the cosmic shadows since the universe's first moment. All delivered in a flash of ghost particles that nearly missed us entirely.