Physicists in Romania have used a beam of artificially produced muons to image a structure for the first time. The result, which builds on last year’s demonstration of the first laser-based muon source, raises hopes that muon-based imaging will become more widely available for structures not suited to traditional X-ray techniques.
Muons are heavier cousins of electrons, and their larger mass allows them to penetrate further into thick, dense objects than is possible with conventional electron tomography. For this reason, muon tomography, or muography, is most often used to image large-scale structures in archaeology, geosciences and civil engineering. It is also used in situations that require non-invasive scanning – for example, imaging the core of a burning blast furnace or a nuclear waste storage facility.
To date, muography has relied on muons that are produced naturally when cosmic rays collide with atoms 15–20 km above the Earth’s surface. At just 1 particle/cm2/s, however, this cosmic muon flux is low, and it only comes from one direction: straight up. The combination of low flux and fixed directionality means that accumulating enough data for tomography can take months.
An artificially produced muon beam would avoid these shortcomings, and in 2025, physicists led by Davide Terzani in the Accelerator Technology and Applied Physics division at the Lawrence Berkeley National Laboratory in the US took an important step towards developing one. Their compact muon source is based on a technique called laser plasma acceleration (LPA) that generates electrons with energies in the 10 GeV range. When these energetic electrons hit a solid target, they produce a shower of secondary particles, including electrons, positrons, gamma rays, neutrons and – importantly – muons.
Designing a shielding structure
In the new work, which is described in an arXiv preprint, teams led by Alexandra Saftoiu and Domenico Doria in the Horia Hulubei National Institute for Physics and Nuclear Engineering used the 10 petawatt (PW)-class laser at the Extreme Light Infrastructure Nuclear Physics (ELI-NP) facility to produce a similar muon beam. Then they set out to optimize it.
“The main challenge here,” Saftoiu explains, “was to achieve an optimal balance between suppressing the background particles, unavoidably coming from the same process that creates muons, to a level that allows us to unambiguously detect muons, while keeping the muon beam high in flux and energy.”
To this end, the researchers used the GEANT4 platform to perform extensive Monte Carlo simulations of how different types of particles pass through matter. The results of these simulations indicated that an optimal, background-suppressing shield would have three layers. The first layer comprises stacks of polyethylene sheets and bricks with a total thickness of 100 cm along the axis of the muon beam. The second layer, which is 60 cm thick along the beam axis and 40 cm beside and above it, consists of paraffin bricks and encapsulates the entire polyethylene structure. An additional layer composed of a lead beam dump and a two-metre reinforced concrete wall completes the shielding.
Once they had built this shield, the teams used three different scintillator-based detectors to monitor the signal from the beam. Measurements from these detectors show that 90% of the particles in the beam were indeed muons. It was this beam that the researchers used, as a proof of concept, to create a low-resolution image depicting the shadow of a pile of lead bricks placed in front of a van.
Good prospects, but a long way to go
Doria says that an artificial muon source would have many applications. Among the most prominent are border security, where it could be used to scan containers for illegally trafficked weapons or radioactive materials, and nuclear safety, where it would lend itself to tasks such as scanning waste-storage casks or even the nuclear reactors themselves. “In principle, it could also be employed in resource management (scanning for ores with a single borehole to probe distances of kilometres rather than just metres), archaeological searches for hidden spaces and structures, safety assessment (for example, scanning dams, tunnels and mine galleries) and even natural hazard assessment (monitoring volcano and glacier activity, for instance),” Doria adds.
Cosmic muons monitor river sediments surrounding Shanghai tunnel
According to Terzani, who was not involved in this work, the ELI-NP study shows progress in developing filters that can isolate muon signals from unwanted background. “Without removing secondary particles, detectors would be swamped by noise that is difficult to subtract, resulting in essentially unusable images,” he says. “This is like trying to photograph a partial solar eclipse without a filter on your camera: even if the Sun is 99% covered, the remaining light will still overwhelm the camera’s sensors.”
Both Saftoiu and Doria emphasize, however, that much work is still needed before laser-driven artificial muon sources can be deployed in real-world situations. “First, the laser technology employed in this context needs to be made substantially more compact and be able to operate at higher repetition rates,” they say. “Muography also requires detectors with higher spatial resolution, capable of measuring intense muon bunches rather than relying on single-particle counting.” Finally, Saftoiu and Doria note that it would be helpful to develop even more effective shielding strategies, with the aim of further minimizing the background generated alongside the muons.
