⚛️ high-energy experiments

Disappearing Tracks from Electroweak Vector Dark Matter at the LHC

This paper analyzes the disappearing-track signature of a spin-1 electroweak vector dark matter candidate at the LHC, using full Run-2 data to exclude quintuplet masses up to 1080 GeV and predict future sensitivity reaching 1750 GeV with the High-Luminosity LHC.

Original authors: Fabián Hernández-Pinto, Patricio Escalona, Alfonso Zerwekh

Published 2026-10-14

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Fabián Hernández-Pinto, Patricio Escalona, Alfonso Zerwekh

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The universe is filled with matter that we cannot see. Astronomers have measured its gravitational pull on galaxies and the cosmic web, confirming that this invisible substance, known as dark matter, makes up most of the mass in the cosmos. Yet, despite decades of searching, no one has directly detected a single particle of dark matter. The leading theory suggests these particles are "weakly interacting massive particles," or WIMPs. These are heavy, invisible entities that rarely bump into ordinary matter, making them incredibly difficult to catch. To find them, physicists build massive detectors deep underground or use powerful particle colliders like the Large Hadron Collider (LHC) in Switzerland. The idea is to smash protons together at nearly the speed of light, hoping to create a dark matter particle that escapes the collision, leaving behind a signature of missing energy or a strange, lingering track in the detector.

A team of researchers recently turned their attention to a specific, elusive candidate for this dark matter: a heavy, invisible particle that comes with a charged partner. In this scenario, the dark matter particle is part of a larger family of particles that interact through the weak nuclear force. While the main dark matter member is neutral and invisible, the family also includes a charged version. Because of a tiny difference in mass between the two, the charged partner is unstable and eventually decays into the invisible dark matter particle. However, this decay does not happen instantly. The charged particle travels a measurable distance inside the detector before vanishing. This creates a unique signal: a short, straight line in the tracking system that simply stops, as if the particle disappeared into thin air. This "disappearing track" is the specific signature the researchers set out to find.

The scientists focused on a theoretical model where the dark matter candidate is a spin-1 particle, meaning it has a specific type of intrinsic angular momentum, and exists in a five-member family. In this model, the charged member of the family is slightly heavier than the neutral dark matter member. This small mass difference is crucial. It forces the charged particle to decay into the neutral one and a very light, soft pion (a type of subatomic particle). Because the energy released is so small, the pion moves too slowly to be detected by the outer layers of the machine. The charged particle, however, travels far enough to leave a clear mark in the innermost layers of the detector before it decays. The result is a track that starts in the center of the detector and abruptly ends, with no corresponding hit in the outer layers.

To test this idea, the researchers analyzed data from the ATLAS experiment at the LHC, using the full dataset collected during the second run of the machine, which corresponds to 137 inverse femtobarns of collisions. They did not look for new data but instead re-examined existing search results. The ATLAS collaboration had previously looked for disappearing tracks to find long-lived particles predicted by supersymmetry, a different theory of physics. The new team adapted this search to look specifically for the spin-1 dark matter candidate. They built a computer simulation of how these particles would behave, how they would be produced in collisions, and how they would move through the detector. They calculated the probability that a particle with a specific lifetime would leave a track that the detector could see, accounting for the fact that the particle must travel far enough to reach the outer layers of the inner tracker but not so far that it reaches the next layer of the detector.

The analysis revealed that the current data from the LHC has already ruled out a significant range of possibilities for this type of dark matter. If the charged particle has a lifetime that allows it to travel about 250 millimeters before decaying, the experiment has excluded the existence of these particles with masses up to 1080 gigaelectronvolts. This is a heavy mass, roughly 1000 times the mass of a proton. For particles with shorter lifetimes, the limit drops to 450 gigaelectronvolts, and for those with longer lifetimes, the limit remains high at around 855 gigaelectronvolts. The researchers also looked at the relationship between the particle's mass and the tiny mass difference that drives the decay. They found that for a specific mass difference of 166 megaelectronvolts, particles with masses below 630 gigaelectronvolts are excluded. If the mass difference is 200 megaelectronvolts, the limit is 395 gigaelectronvolts.

Perhaps most importantly, the team checked their findings against the specific predictions of their theoretical model. The model predicts that the mass difference between the charged and neutral particles is generated by quantum effects and depends on how the particles interact with the Higgs field. For a specific, small interaction strength, the model predicts that particles with masses between 120 and 300 gigaelectronvolts should exist. However, the experimental data shows that particles in this entire mass range, with the predicted mass differences, do not exist. The model is effectively ruled out for these masses. The researchers noted that if the interaction strength were larger, the predicted mass difference would be bigger, causing the particles to decay too quickly to be seen by this specific search, leaving that part of the theory unconstrained.

Looking ahead, the researchers calculated what would happen if the LHC collected three times more data, a scenario expected in the future. With this larger dataset, the search would become sensitive to much heavier particles. If the background noise from random detector errors could be kept to zero, the experiment could potentially find or rule out particles with masses up to 1750 gigaelectronvolts. Even with a more realistic estimate of background noise, the reach would extend to 1530 gigaelectronvolts. This would cover a vast range of masses that are currently unknown. The study demonstrates that the disappearing track signature is a powerful tool for hunting down these heavy, invisible particles. While this specific model of dark matter is excluded in the lower mass range, the method remains a vital strategy for exploring the unknown corners of the universe. The search continues, with the LHC ready to peer deeper into the subatomic world than ever before.

Technical Summary: Disappearing Tracks from Electroweak Vector Dark Matter at the LHC

Problem and Motivation
The Standard Model (SM) fails to account for dark matter (DM), necessitating extensions such as Weakly Interacting Massive Particles (WIMPs). While scalar and fermionic minimal dark matter (MDM) candidates have been extensively studied, spin-1 realizations remain relatively unexplored due to theoretical challenges regarding perturbative unitarity for massive vector fields. This paper investigates a specific spin-1 MDM candidate: a real five-dimensional representation (quintuplet) of $SU(2)_L$ with vanishing hypercharge ( $Y=0$ ).

In this model, radiative electroweak corrections lift the tree-level mass degeneracy among the multiplet components, creating a compressed spectrum. The neutral component ( $V^0$ ) is stable and serves as the DM candidate, while the charged components ( $V^\pm$ ) acquire a small mass splitting ( $\Delta m$ ) relative to $V^0$ . This splitting, typically in the sub-GeV range, suppresses the decay width of the charged states, allowing them to traverse macroscopic distances within the detector before decaying via $V^\pm \to V^0 \pi^\pm$ . This results in a "disappearing track" signature—a short track segment in the inner detector that terminates without hits in the outer tracking layers. The authors aim to constrain this model using existing LHC data, specifically addressing the need for a dedicated reinterpretation of searches originally designed for supersymmetric charginos, given the distinct spin and production kinematics of the vector quintuplet.

Methodology
The study employs a recasting strategy of the ATLAS disappearing-track search using the full Run-2 dataset ( $137 \text{ fb}^{-1}$ at $\sqrt{s} = 13 \text{ TeV}$ ).

Model Implementation: The authors utilize a UFO model generated with FeynRules, incorporating the effective Lagrangian for the vector quintuplet. To ensure perturbative unitarity, non-minimal gauge couplings ( $\kappa$ ) and specific self-interaction parameters ( $\alpha_2$ ) are set to zero, while the Higgs-portal coupling ( $\lambda_{HV}$ ) and quartic vector interactions are constrained. The mass splitting $\Delta m$ is computed at one-loop order using FeynArts and FeynCalc, accounting for renormalization scale uncertainties.
Event Generation: Signal events are generated at leading order (LO) using MadGraph5_aMC@NLO with Pythia 8 for parton showering and hadronization. The simulation includes all relevant pair-production channels ( $pp \to V^\pm V^0$ , $V^+ V^-$ , $V^{\pm\pm} V^\mp$ , etc.) with initial-state radiation (ISR) to provide the necessary missing transverse momentum ( $p_T^{miss}$ ) and boost.
Detector Emulation and Selection: A simplified detector emulation based on the TOYDETECTOR framework mimics the ATLAS geometry. The analysis focuses on the SR4High signal region, which requires a four-layer tracklet in the Pixel Detector and zero associated hits in the Semiconductor Tracker (SCT). Selection criteria include high- $p_T$ jets, large $p_T^{miss}$ , and specific isolation requirements for the tracklet.
Efficiency Calculation: A key methodological contribution is the analytical computation of the tracklet reconstruction probability. Instead of regenerating events for every lifetime, the authors integrate the exponential decay distribution of the charged particle over the published ATLAS efficiency function as a function of the decay radius. This allows a single event sample per mass point to cover the full range of proper decay lengths ( $c\tau$ ).
Validation: The procedure is validated against the ATLAS limits for pure winos (a fermionic benchmark). The recast reproduces the ATLAS mass limits within 6%, confirming the reliability of the analytical efficiency approach.

Key Contributions and Results
The paper provides the first comprehensive constraints on the spin-1 vector quintuplet DM model using disappearing-track signatures.

Constraints in the $(M_V, c\tau)$ Plane: Using Run-2 data, the authors exclude quintuplet masses ( $M_V$ ) up to 1080 GeV for proper decay lengths around $c\tau \simeq 250 \text{ mm}$ . The exclusion limit remains above 1 TeV for $0.1 \text{ m} \lesssim c\tau \lesssim 1 \text{ m}$ . These limits are stronger than those for winos at comparable lifetimes due to the larger production cross-section of the vector quintuplet.
Constraints in the $(M_V, \Delta m)$ Plane: Translating the lifetime limits into mass splitting constraints, masses below 630 GeV are excluded for $\Delta m = 166 \text{ MeV}$ , and masses below 395 GeV for $\Delta m = 200 \text{ MeV}$ .
Model-Specific Exclusions: For the benchmark scenario with a small Higgs-portal coupling ( $\lambda_{HV} = 0.01$ ), the one-loop radiative mass splitting prediction is excluded for masses between 120 GeV and 240–300 GeV (depending on the renormalization scale choice). In this range, the predicted splitting falls between the charged pion and $\rho$ -meson thresholds, maximizing the disappearing track signature.
Future Projections: Projections for the High-Luminosity LHC (3 ab $^{-1}$ ) suggest a maximum reach of 1750 GeV (assuming vanishing background) or 1530 GeV (assuming a linear extrapolation of instrumental background).

Significance and Claims
The authors claim that their work establishes the viability of the disappearing-track search as a powerful probe for spin-1 dark matter candidates. By demonstrating that the vector quintuplet can be constrained up to the TeV scale, they highlight that collider searches remain a critical complementary probe to direct and indirect detection, particularly for models where the DM-nucleon scattering cross-section is suppressed (e.g., via the absence of tree-level $Z$ -exchange).

The paper modestly notes that its results rely on a single-bin counting experiment, LO signal normalization, and a simplified detector emulation. It explicitly states that the analysis is restricted to the regime $\Delta m > m_\pi$ , where the two-body pion decay dominates. For larger masses where the splitting falls below the pion mass, the charged state would decay leptonically with macroscopic lifetimes (meters), a regime outside the scope of this specific study and requiring dedicated future investigations. The authors conclude that a dedicated experimental analysis with upgraded detectors and data-driven backgrounds is necessary to fully cover the parameter space of the vector quintuplet model.

Technical Summary: Disappearing Tracks from Electroweak Vector Dark Matter at the LHC

More like this