Relativistic plasma fireballs created at CERN have opened an unexpected window into a part of the Universe that rarely reveals itself. A beam of electrons and positrons, born inside a dense converter struck by 440 GeV protons, was forced to drift through a column of argon plasma just one meter long. That short journey produced a result that unsettles one of the longest standing puzzles in high energy astrophysics. Observers have watched blazars release TeV gamma rays for years. Those gamma rays should collide with faint background light across intergalactic space, split into torrents of electron and positron pairs, and then scatter cosmic microwave photons into the GeV band. That scattered glow has been searched for repeatedly. It should be bright, extended, and unmistakable. Instead it is gone. Something suppresses it or bends it away from Earth. For more than a decade, two explanations have competed quietly in the background of the field. One suggests that vast magnetic structures hidden between galaxies redirect the pairs. The other claims that the pairs never make it that far. According to that idea the beams tear themselves apart through plasma instabilities, losing coherence and energy long before they can radiate.
The mystery has lingered because the physics is unreachable in ordinary environments. The beams involved in blazar cascades are thin, fast, and nearly neutral. No laboratory had ever generated anything similar. The new CERN experiment built for this purpose approaches the problem from a different angle. Rather than trying to copy the Universe, the team recreated its governing conditions. A graphite and tantalum target was placed in the path of the Super Proton Synchrotron beam. When the protons strike the material they generate neutral pions that decay instantly into gamma rays. The gamma rays initiate a cascade inside the tantalum converter. More than ten trillion electrons and positrons emerge. They form a compact fireball with a realistic mixture of energies and angles. This is not a simplified or ideal beam. It carries the same disorder that exists in the distant regions where blazar photons are converted into matter.
The team then directs this fireball into a plasma column built to behave like the tenuous medium of cosmic voids. The argon plasma is nearly collisionless. Its density and temperature are measured by Langmuir probes before each run and monitored by optical emission spectroscopy. A set of inductive coils maintains the discharge with stable density peaks near the center. The plasma has a skin depth of a few millimeters, matching the scale on which electromagnetic instabilities would form. If the pairs were prone to break into filaments, the plasma would provide the conditions for that growth. To detect even the faintest signs of instability, the team suspended a terbium gallium garnet crystal inside the plasma stream and passed a polarized laser through it. Any magnetic field aligned with the crystal would rotate the laser’s polarization. Farther downstream, an alumina screen would flash on impact with charged particles, revealing the beam’s shape after its passage through the plasma. Between those two diagnostics the entire story of the beam’s evolution could be seen.
The story turned out to be quiet. The Faraday probe registered fluctuations at the level of electronic noise. Combining several exposures sharpened the result. The upper limit of the magnetic field generated inside the plasma is on the order of a few millitesla. For an instability to redirect or break the beam, the field would need to grow far beyond that limit. The screen images were no more revealing. The beam that passed through the plasma looked like the beam that had not. The shapes matched, the widths matched, and no evidence of filamentation appeared. Something that was expected in ideal theoretical systems did not appear in the real environment of the experiment.
The absence of obvious structure might be dismissed as an artifact of the probe, but the researchers anticipated that concern. They repeated the experiment under different plasma pressures and with the probe rotated to detect radial fields. The results remained unchanged. To push further they ran large scale three dimensional particle in cell simulations capable of tracking the motion of electrons, positrons, and residual protons through the plasma. These simulations used the real distributions measured by FLUKA models of the target. When the pairs were artificially forced to be perfectly collimated and monoenergetic, the instability ignited immediately. Filaments formed, then separated, and magnetic fields grew to tens of millitesla. But when the pairs were given the same divergent momenta produced in the experiment, the instability lost its foothold. The angular spread carried particles across the growing modes too quickly for the fluctuations to take hold. The simulated magnetic field stayed at the same weak level found in the laboratory. The growth rate fell far below its ideal value. The physics of the real beam, with all of its natural disorder, acted like a solvent that dissolved the instability before it could form.
Once this behavior was established, the team scaled the results to intergalactic space. The density of pairs produced by a blazar is extremely low. The background plasma of cosmic voids is even thinner. Under these conditions the growth rate of any electromagnetic instability becomes vanishingly small. The preprint calculates that the timescale required to grow even a marginal field is longer than the time available before the pairs lose energy by scattering cosmic microwave photons. The fields that would form after saturation are calculated to be smaller than ten to the minus twenty four tesla. At that scale the pairs would feel nothing. They would travel as if the instability were not there.
This comparison does not require any assumption about distant environments. It follows from the measured suppression in the laboratory. The experimental bound on the growth rate becomes a hard limit when translated to the voids of the Universe. The idea that the missing GeV light is erased by plasma processes is no longer viable. The beam does not disrupt itself. The pairs remain coherent. Their journey through space does not end in a turbulent collapse. Yet the GeV light is missing. Something else has turned it aside.
The possibility left standing is the presence of magnetic fields in intergalactic space. These fields would not resemble the familiar arrangements of planets or stars. They would stretch over regions that contain almost no matter and no visible structure. Their strength would be low by human standards, yet their influence would accumulate over the enormous distances between galaxies. A pair created far from Earth would travel a vast distance before scattering a microwave photon. During that time even a faint field would bend its path. The scattered photon would arrive far from the blazar’s line of sight. Instruments that search for concentrated emission would find nothing. The light would be lost not because it was never produced but because it arrived everywhere.
If such fields exist, their origin does not come from the late history of the cosmos. There is not enough matter in voids to generate them through local processes. Simulations of cosmic structure formation do not produce them spontaneously. Their scale and coherence suggest something older. The preprint notes that the required strength and length make conventional astrophysical explanations unlikely. That leaves the early epochs of the Universe as the only stage large enough to set them in place. Various ideas appear in the literature, from magnetic seeds produced during primordial transitions to fields driven by particles that no longer exist. None of those ideas can be tested directly, but their fingerprints would linger in the behavior of blazar light crossing the void.
The CERN experiment does not attempt to trace those origins. It reveals a boundary. Plasma instabilities cannot erase the cascade. The blazar beam survives, and the missing emission must be redirected by something that occupies the gaps between galaxies. That presence is not directly observed. It is detected only by its influence on what fails to arrive. The fireball experiment creates a scaled version of that journey and shows what does not happen inside it. The beam does not fragment. The field does not rise. The cascade does not collapse.
With this result the puzzle around the missing light becomes more pointed. A process once considered a plausible alternative has been tested in the only environment where it can be observed in detail. The instability faltered. The plasma remained quiet. The beam crossed the chamber almost untouched. The voids of the Universe are far larger and far more rarefied, but the same physics applies. The pairs created by blazars travel through a medium where this instability cannot grow. Yet their expected glow is gone.
This silence is not empty. It hints at structures that cannot be seen directly but shape the paths of particles across cosmic distances. The experiment at CERN does not illuminate those structures. It only reveals that the plasma between galaxies is not the thing that hides the light. Something older and more elusive stands between the blazars and Earth. The missing glow from the cascade is not a failure of the pairs themselves. It reflects a landscape that remains hidden even as its effects pass through every telescope searching for the faintest trace of the Universe’s highest energy particles.
Source:
https://arxiv.org/abs/2509.09040
Arrowsmith, Charles D., et al. “Suppression of Pair Beam Instabilities in a Laboratory Analogue of Blazar Pair Cascades.” Proceedings of the National Academy of Sciences, vol. 122, no. 45, 2025, article e2513365122. https://doi.org/10.1073/pnas.2513365122.
