First demonstration of $50\,\mathrm{ps}$ four-dimensional vertex reconstruction with the Timepix4 beam telescope

In the world of high-energy physics, scientists smash particles together at incredible speeds to study the fundamental building blocks of the universe. These collisions happen so frequently that, within a single instant, dozens of separate events can occur at the exact same location. Imagine trying to sort through a pile of letters that were all dropped on the floor at the same time; if you only look at where they landed, you cannot tell which letter belongs to which envelope. For decades, researchers have relied on the position of these particles to sort them out. However, as machines become more powerful, the collisions happen so close together in space that position alone is no longer enough. To solve this, scientists are turning to time. If they can measure exactly when a particle arrives with extreme precision, they can separate events that happen at the same spot but at slightly different moments. The goal is to reach a level of timing so sharp that it can distinguish between collisions separated by just a few tens of picoseconds, a timescale so brief that light travels only a few millimeters in that duration.

A team of researchers has now taken a major step toward this goal by demonstrating a new way to reconstruct the origin of particle collisions using both space and time. Working with a specialized telescope made of silicon sensors at the European Organization for Nuclear Research, they successfully tracked particles produced by a high-speed beam hitting a fixed target. By combining measurements from multiple layers of detectors, they were able to pinpoint not just where a collision happened, but also exactly when it occurred, achieving a time resolution of 54 picoseconds. This is the first time such a system has been shown to work effectively for reconstructing the full four-dimensional history of a collision point, proving that time can be used as a powerful tool to untangle crowded particle events.

The experiment took place at a facility in Switzerland where a beam of mixed hadrons, a type of subatomic particle, was fired at a stationary target. As the beam struck the target, it created a shower of new particles flying outward. To catch these particles, the team used a device called a beam telescope, which consists of eight flat layers of silicon sensors arranged in two arms. Each layer is covered in tiny pixels that act like digital eyes, recording not only where a particle passed through but also the precise moment it arrived. The researchers directed the beam at a target made of tungsten and brass, creating a controlled environment where they could study how well their system could determine the starting point of the particle showers. They collected data over several months, capturing millions of these interactions to build a detailed picture of how the particles moved and when they arrived.

To make sense of the data, the researchers developed a new method of analysis that treats time as a fundamental part of the reconstruction process, rather than just an afterthought. In previous approaches, scientists would first figure out the path a particle took through space and then try to add time information later. In this study, the team fitted the path and the time of arrival simultaneously, allowing the time measurements to help define the path just as the spatial measurements helped define the time. This approach required sophisticated software to handle the complex math of combining data from different layers, especially since the particles were flying at wide angles and hitting the sensors in unpredictable ways. The system also relied on a separate timing reference, using two specialized detectors to provide a master clock for every event, ensuring that the measurements from the telescope were synchronized with extreme accuracy.

The results showed that the system could determine the location of a collision with impressive precision. For the position along the direction of the beam, the resolution improved as more particle tracks were included in the calculation, reaching a precision of about 0.66 millimeters when six tracks were used. For the position across the beam, the system achieved a resolution of about 20 micrometers with five tracks. These spatial measurements confirmed that the telescope could accurately map the geometry of the collision. More importantly, the team measured the time resolution of the reconstructed collision point. As they combined more timing measurements from the sensor layers, the precision of the time measurement improved, following the expected statistical pattern. When they used data from 16 timing clusters, the system reached a time resolution of 54 picoseconds.

This achievement is significant because it proves that there is no hidden barrier preventing the system from getting even better. The researchers found that the improvement in time resolution followed the expected trend, meaning that adding more measurements continued to sharpen the picture without hitting a sudden wall of inaccuracy. However, they also identified a small, unchangeable limit in their system. Even with perfect measurements, a residual uncertainty of about 29 picoseconds remained. This limit was not caused by the sensors themselves or by the way the particles were measured, but rather by the clock system used to synchronize the data and the physical layout of the timing reference detectors. The researchers determined that the time it took for light to travel through the reference detectors varied slightly depending on where the particles hit, introducing a small amount of uncertainty that could not be eliminated by simply adding more sensors.

Despite this small limitation, the experiment marks a clear milestone. It demonstrates that silicon pixel detectors, which are widely used in particle physics, can be combined to achieve four-dimensional reconstruction at the 50-picosecond scale. This capability is essential for the future of high-energy physics, where machines will produce collisions so dense that spatial information alone will fail to separate them. By showing that time can be used to sort these events with high precision, the study provides a roadmap for the next generation of detectors. The researchers concluded that while their current system did not yet reach the ultimate goal of 20 picoseconds required for future upgrades, it proved that the path forward is clear and that the technology is ready to be refined for even greater precision.