Strategy and performance of the CMS long-lived particle trigger program in proton-proton collisions at $\sqrt{s}$ = 13.6 TeV

The universe is built on a set of rules known as the standard model, a theory that successfully describes how the smallest known particles interact. Yet, this theory leaves major questions unanswered, such as the nature of dark matter or why there is more matter than antimatter. Physicists suspect that the missing pieces of the puzzle might be hidden in a class of particles that do not behave like the ones we see every day. These hypothetical particles are called long-lived particles. Unlike ordinary particles that vanish almost instantly after being created, long-lived particles can travel a measurable distance through a detector before they decay or disappear. They might move slowly, decay far from the point where they were born, or leave behind strange, delayed signals that standard equipment is not designed to catch. Finding them is crucial because they could be the key to understanding the dark matter that holds galaxies together.

To hunt for these elusive travelers, the CMS experiment at the Large Hadron Collider in Switzerland smashes protons together at tremendous speeds. The resulting collisions create a shower of particles that fly out in all directions, passing through a massive, multi-layered detector. The challenge is that the detector sees millions of collisions every second, and it cannot save data for every single one. It must make split-second decisions to keep only the most interesting events. For years, the computer systems that make these decisions were programmed to look for particles that appear immediately at the collision point. This meant that if a new, strange particle traveled a few centimeters or even several meters before decaying, the trigger system would likely ignore it, assuming it was just background noise. The paper presented here details a major upgrade to this decision-making system, specifically designed to catch these long-lived particles during the latest run of the collider.

The researchers describe a comprehensive program of new digital filters, or triggers, that were introduced to the CMS detector between 2022 and 2024. These new tools allow the experiment to look for signatures that were previously invisible. Instead of just waiting for a particle to appear right at the center of the collision, the new triggers can spot a particle that appears later, further away, or with a different timing than expected. The team tested these systems using data collected during 2022–2024, corresponding to integrated luminosities of up to 123 fb−1, a volume of information that required the detector to run at its highest capacity. They found that these new strategies successfully identified events that would have been discarded by the old system, effectively widening the net for new physics.

One of the most significant changes involves how the detector tracks particles that seem to vanish. In some scenarios, a charged particle might travel a short distance inside the tracking chamber and then decay into something the detector cannot see. To the computer, this looks like a track that simply stops in mid-air. The new triggers are programmed to look for these "disappearing tracks" by checking if a high-energy particle suddenly has no hits in the outer layers of the detector. Another new strategy focuses on particles that arrive late. Because light travels at a constant speed, a heavy, slow-moving particle will take longer to reach the outer sensors than a fast-moving one. The upgraded system can now measure the precise arrival time of energy deposits in the calorimeters, the parts of the detector that absorb particle energy. If a jet of particles arrives a few billionths of a second later than expected, the new triggers flag it as a potential long-lived particle.

The paper also details how the experiment looks for particles that decay inside the muon system, the outermost layer of the detector. Normally, muons pass through the detector in a straight line, leaving a clean trail of hits. However, a heavy, long-lived particle might decay inside this outer layer, creating a chaotic shower of hundreds of hits in a small area. The new triggers are designed to recognize this specific pattern of high-multiplicity clusters, distinguishing them from the clean trails of ordinary muons. Additionally, the team developed methods to look for particles that decay when the proton beams are not even colliding. These "stopped" particles might travel through the detector, come to a halt in the dense metal shielding, and decay in the empty time windows between beam trains, specifically in bunch crossings at least two intervals away from any proton collisions. Special triggers activate during these quiet moments to catch these delayed signals.

The performance of these new tools was tested against a variety of theoretical models, including those involving supersymmetry and hidden sectors of physics. The results show that the new triggers are highly effective. For example, the system can now detect jets of particles that originate from a decay point several meters away from the collision center, a region that was previously difficult to access. The efficiency of finding these events has improved dramatically compared to the previous run of the collider, with some new triggers offering up to ten times better sensitivity for certain types of signals. The researchers also demonstrated that these triggers can work together, covering different distances and types of decays to ensure that no potential signal is missed.

A key part of this success is the ability to handle the immense data rate of the collider. The new triggers are sophisticated enough to filter out the overwhelming background of ordinary collisions while keeping the rare, strange events. The team used a technique called "scouting," where they save only the most important summary information about an event rather than the raw data, allowing them to keep a much higher rate of interesting events. This approach has already led to the publication of new searches for displaced dimuons and displaced jets, which show sensitivity comparable to or better than previous searches, despite using a fraction of the data.

The paper concludes by outlining how these improvements lay the groundwork for the future. As the collider moves toward even higher energies and luminosities in the coming years, the ability to spot long-lived particles will become even more critical. The new algorithms and hardware upgrades described here have already proven their worth, turning the CMS detector into a more versatile instrument capable of seeing the invisible. By expanding the range of what the machine can observe, the researchers have opened a new window into the fundamental laws of nature, bringing us closer to understanding the hidden sectors of the universe that have so far remained out of reach.