Some breakthroughs in physics come from brand new inventions. Others begin with a new theory. But many advances happen when researchers combine familiar technologies in an unexpected way and create something more powerful than the individual parts.
That strategy could be especially valuable in the search for weakly interacting particles, including neutrinos and certain dark matter candidates. These particles are notoriously difficult to detect because they rarely interact with ordinary matter. Building larger detectors and improving their spatial resolution can increase the odds of observing the faint signals they produce, but doing so often makes the instruments more complicated and expensive.
Similar demands apply to calorimeters, the devices used in collider experiments to measure the energy carried by particles.
Why Particle Detectors Are So Complex
Most particle physics experiments need to reconstruct the three-dimensional (3D) paths of elementary particles as they move through large volumes of dense material.
One common detector material is a scintillator. When a charged particle passes through a scintillator, the material gives off tiny flashes of visible light. Scientists use those flashes to determine where the particle traveled and how it interacted with the detector.
To pinpoint the particle’s location, the scintillator is usually divided into a vast number of small active sections. Optical fibers collect the photons produced in each section and carry the light to photomultiplier tubes or silicon photomultipliers, which count the photons.
This approach can be highly precise, but it becomes difficult to scale.
The T2K neutrino-oscillation experiment in Japan, for example, uses a detector with about two tons of sensitive material made from approximately two million cubes and 60,000 fibers. At CERN and the Paul Scherrer Institute, the LHCb and Mu3e experiments reach sub-millimeter spatial resolution by using millions of thin scintillating optical fibers.
These systems demonstrate what segmented detectors can accomplish, but they also reveal a growing problem. As detectors become larger, manufacturing, assembling, and reading out millions of individual components can become a major technological and financial bottleneck.
A Radical New Approach to Particle Tracking
Researchers at ETH Zurich and EPFL are now proposing a very different strategy.
PhD student Till Dieminger, senior scientist Dr. Saúl Alonso-Monsalve, Professor Davide Sgalaberna and colleagues in his group, together with members of the Advanced Quantum Architecture Lab at EPFL in Lausanne led by Professor Edoardo Charbon, developed and tested the first prototype of a detector designed to perform ultrafast, high-resolution 3D particle imaging inside a large, unsegmented block of scintillator material.
Instead of dividing the detector into millions of tiny units, the system uses advanced camera technology to reconstruct where the light originated.
The prototype demonstration and an extensive series of simulations were described recently in Nature Communications.
Turning Light Field Photography Into a Physics Tool
The detector draws inspiration from plenoptic cameras, also known as light field cameras.
Unlike an ordinary camera, which mainly records the intensity of incoming light, a light field camera also captures information about the direction from which the light arrived. This allows it to recover depth and reconstruct a scene in three dimensions.
The technology relies on a micro-lens array (MLA) placed between the camera’s main lens and imaging sensor. Each microscopic lens acts like a tiny camera, recording the same scene from a slightly different angle. When the information from all of these lenses is combined, the system can reconstruct a light field, which describes the intensity, position, and direction of the incoming light.
For particle detection, this ability is particularly useful because the light inside a scintillator may be extremely faint.
When plenoptic cameras are paired with single-photon avalanche diode (SPAD) array sensors, they can detect individual photons and potentially reconstruct particle tracks even when very little light is available. Despite that promise, light field cameras had not previously been explored for particle tracking.
Inside the PLATON Prototype
The new system was developed through the PLATON project, which is funded by the Swiss National Science Foundation.
The ETHZ-EPFL team built a proof-of-concept detector that combines a micro-lens array with a SPAD imaging sensor. The sensor, known as SwissSPAD2, was developed by the EPFL team. Raytrix GmbH designed the MLA and mounted it directly onto the sensor to create the complete plenoptic imaging system.
SwissSPAD2 also provides gated photon detection. This means that the sensor records photons only within defined time windows.
That timing control helps researchers focus on periods when genuine scintillation light is most likely to be present while filtering out random background signals and other spurious counts.
Testing the Detector With Just a Few Photons
The researchers tested PLATON’s spatial resolution in laboratory experiments using light levels ranging from several hundred detected photons down to only five.
They also evaluated whether the prototype could detect electrons and reconstruct their positions inside a block of plastic scintillator. The electrons were produced using a strontium-90 source.
Across the different test conditions, the simulations closely matched the laboratory measurements, giving the researchers confidence that their models accurately describe the detector’s performance.
The results from the first demonstrator have already shaped the team’s plans for the next version of PLATON.
Faster Timing and Greater Sensitivity
The researchers are developing a new SPAD array sensor designed to improve photon detection efficiency and provide sub-nanosecond timing for individual photons.
In the current system, photons are assigned to fixed time windows. In the upgraded version, each detected photon would receive its own precise time stamp.
That added timing information could help the system determine more accurately where each photon came from and improve the reconstruction of particle tracks.
The researchers have also optimized the plenoptic camera to expand its field of view and collect more light. Simulations presented in the paper suggest that these changes should further improve PLATON’s spatial resolution.
AI Reconstructs Hidden Particle Interactions
The team also used simulations to estimate how an upgraded PLATON system could perform when detecting neutrinos.
The simulations incorporated a new image-processing method based on a neural network (NN). The system uses a Transformer architecture adapted from the type commonly used in large language models.
Rather than analyzing words, however, this Transformer examines patterns among the scintillation photons recorded by the detector. It is designed to identify correlations in where and when the photons appear, allowing it to reconstruct the original particle interaction.
The simulations indicate that an unsegmented PLATON detector with a volume of (10x10x10)cm3 could realistically achieve spatial resolution below 1mm.
They also suggest that the system could identify neutrino interactions that produce final-state low-momentum protons with both high purity and high efficiency. In other words, the detector may be able to select the desired events while rejecting many unrelated signals.
Scaling Up to a Cubic Meter
The researchers also considered how the technology might perform in a much larger detector.
Because of limited computing resources, they did not run full neutrino simulations for a one-cubic-meter block of unsegmented scintillator. Instead, they modeled a simplified point-like source of photons.
The simulations suggest that a detector of this size could achieve spatial resolution of a few millimeters, placing it on par with state-of-the-art plastic scintillator detectors.
The result is especially notable because PLATON would achieve this performance without dividing the scintillator into millions of individual pieces.
The authors believe that additional improvements to the optical design and other parts of the system could eventually make sub-millimeter resolution possible in PLATON-type detectors with volumes larger than 1m3.
Potential Uses Beyond Particle Physics
The ETH Zurich researchers believe the technology could eventually be useful far beyond neutrino experiments and particle colliders.
Because PLATON is designed to reconstruct the position of faint light signals in three dimensions, it could improve a wide range of imaging systems.
Dieminger, Alonso-Monsalve and Sgalaberna have already filed three separate patents involving the use of PLATON technology in positron emission tomography (PET). PET is a medical imaging method that tracks radioactive tracers inside the body to reveal activity in organs and tissues.
The patents cover both the scanner design and the image-processing techniques, including the NN developed by Alonso-Monsalve.
Particle physics has a long history of producing technologies that later find broader uses. The world wide web was created at CERN, while proton therapy grew from advances in particle accelerators and radiation physics.
PLATON could become another example of a physics experiment leading to a technology with major scientific and medical applications.