LLPs in BSM Searches

• LLPs are particles with macroscopic lifetimes that arise from suppressed decay widths in BSM models, playing a key role in dark matter, baryogenesis, and neutrino mass studies.
• They are produced via exotic processes like Higgs decays, heavy resonance decays, and meson decays, with collider kinematics crucially affecting their detectability.
• Advanced detection strategies—including displaced vertices, timing-based methods, and dedicated detectors (e.g., MATHUSLA, ANUBIS)—extend experimental reach into previously unexplored lifetime regimes.

Long-lived particles (LLPs) are defined as particles whose lifetimes are macroscopic—far exceeding those typical of prompt decay processes—such that their decay positions or times are experimentally resolvable from their production point. LLPs are ubiquitous in extensions of the Standard Model (SM) targeting open problems such as the nature of dark matter, baryogenesis, or the origin of neutrino masses. In collider environments, especially at the LHC and future high-energy facilities, LLPs can have lifetimes ranging from fractions of a millimeter up to thousands of meters, generating non-prompt experimental signatures that test the limits of standard reconstruction tools and trigger strategies (Jeanty et al., 22 Nov 2025). This article presents a comprehensive analysis of LLP origin, detection methodologies, challenges, experimental reach, and the evolving landscape of LLP searches.

1. Theoretical Origin and Physical Parameters

LLPs arise whenever the particle decay width, $\Gamma$, is suppressed such that the mean lifetime $\tau = \hbar/\Gamma$ becomes macroscopic compared to detector scales. Mechanisms effecting this suppression include:

• Heavy mediators: Decays via virtual heavy states lead to weak effective interactions, consequently long lifetimes.
• Small coupling strengths: Portal or mixing parameters can be very small, inducing feeble interactions with SM states.
• Phase-space limitations: Decay energy release may be minimal, slowing down the process.

Explicit examples include heavy neutral leptons with suppressed mixings ($\theta^2 \sim 10^{-10}$–$10^{-6}$), gauge-mediated SUSY NLSPs ($c\tau$ scaling with the SUSY breaking scale $F$), and hidden-valley sectors with confining dynamics yielding long-lived bound states (Jeanty et al., 22 Nov 2025). In the SM, particles such as $K^0_L$, $n$, and $\mu$ are LLPs by virtue of weak or phase-space-limited decays.

2. Production Channels and Collider Kinematics

LLPs can be produced singly or in pairs, either as primary products or through cascade decays, with typical channels encompassing:

• Exotic Higgs decays: $pp \to h \to XX$, where $X$ is an LLP. The cross section scales as $\sigma(pp\to h)\mathrm{BR}(h\to XX)$, with total rates at the HL-LHC exceeding $10^8$ Higgs bosons (Collaboration, 30 Oct 2025).
• Heavy resonance decays: $Z' \to XX$ and $W' \to \ell X$ are prominent in many BSM scenarios.
• Cascade decays: SUSY and Hidden Valley models frequently yield multi-step processes involving long-lived neutralinos or dark mesons.
• Meson decays at low-energy or beam-dump experiments: Light LLPs are often copiously produced in meson decays ($K$, $D$, $B$), enabling high-intensity flavor facilities or neutrino experiments to probe LLP scenarios (Beltrán et al., 2023).

Collider-produced LLPs typically have moderate boosts ($\gamma \sim 1$–few), crucially affecting the laboratory decay length $d = \beta\gamma c\tau$ and the fraction decaying within detector volumes.

3. Experimental Signatures and Detection Strategies

LLP searches exploit non-prompt event topology, focusing on:

• Displaced vertices (DV): Two or more charged tracks reconstructing a vertex far ($r\gg\text{mm}$) from the primary collision; key in tracker, calorimeter, and muon systems. Vertex mass and track multiplicity selection enhance signal-to-background (Gonski, 2022).
• Displaced leptons/jets: Individual displaced objects with anomalously large impact parameter $d_0\gtrsim\text{mm–cm}$.
• Stopped-particle searches: LLPs decelerated and stopped in dense detector regions, decaying out-of-time with collisions (Jackson, 2011).
• Timing-based strategies: LLP decay products arrive at detector modules with nanosecond-scale delays, resolvable with precision timing systems ($\sim$30 ps in MTD, $\sim$1 ns in muon detectors), dramatically suppressing prompt backgrounds (Liu et al., 2018).
• Heavy stable charged particle (HSCP) searches: LLPs traversing the entire detector, identified via anomalous d$E$/dx and time-of-flight measurements.
• Missing energy: LLP decays outside the instrumented detector contribute to non-reconstructable momentum, detected as enhanced $E_T^{miss}$.

Recent deployments incorporate machine learning for event categorization and background rejection, as in future lepton colliders where deep neural network approaches reach signal efficiencies of up to 99% for $m_X=50$ GeV, $\tau\sim1$ ns (Zhang et al., 2024).

4. Dedicated LLP Detectors and Extended Reach

Standard collider detectors struggle to cover the full LLP lifetime regime, motivating dedicated detector proposals:

• ANUBIS: An air-filled decay volume ($R \sim 9.5$–21 m, $|\eta|<0.7$, ceiling location $z=23$ m above ATLAS IP) with tracking-station arrays of RPCs ($99.9\%$ efficiency per layer). ANUBIS probes exotic $h\to XX$ branching ratios down to $\mathcal{O}(10^{-6})$ for $c\tau\sim0.1$–$10^4$ m, surpassing other transverse detectors in both lifetime and branching sensitivity (Collaboration, 30 Oct 2025).
• MATHUSLA: 200 m $\times$ 200 m $\times$ 20 m surface detector with five RPC layers and extensive cosmic-ray vetoing, optimal for c$\tau$ up to the BBN upper limit ($10^7$ m), achieving $\mathrm{BR}(h\to XX)\sim10^{-5}$ for c$\tau\sim200$ m (Chou et al., 2016, Curtin et al., 2018).
• FASER and CODEX-b: Compact forward/transverse detectors focusing on low-mass, highly-boosted LLPs; e.g., FASER-2 at HL-LHC can test Type-I 2HDM scalar/pseudoscalar LLPs for $\tan\beta$ up to $10^6$–$10^7$ (Qi et al., 17 Aug 2025).
• DUNE Near Detectors: Light LLPs produced in meson decays ($K$, $D$) are detectable via background-free channels ($\mu^+\mu^-$, $3\pi$) in the ND-GAr, setting world-leading limits on ALP couplings and lifetimes (Coloma et al., 2023).

These facilities complement traditional detector reach and allow robust coverage of lifetimes from centimeters to kilometers and masses from MeV to multi-TeV.

5. Data-driven Background Estimation and Analytical Approaches

Simulation-based background estimation is often unreliable for rare or detector-specific processes; thus, data-driven techniques are adopted:

• ABCD methods: Orthogonal signal/control triggers (e.g., isolated/non-isolated Muon-RoI clusters) and additional event-split variables $Y$ (e.g., $\Delta\phi(\text{MET},\text{DV})$) enable robust differential background measurement and subtraction in LLP searches (Coccaro et al., 2016).
• Width-ratio reinterpretation: For LLPs produced from similar parent mesons with long travel distance relative to detector location, experimental exclusions on one LLP model can be mapped onto another via theoretical decay-width ratios, eliminating the need for full simulation (Beltrán et al., 2023).
• Machine-learning for clustering and event categorization: Neural networks with image or graph inputs vastly enhance DV reconstruction, particularly at lepton colliders, and can achieve background-free operation over wide $(m_X, \tau)$ ranges (Zhang et al., 2024).

6. Experimental Reach, Benchmarks, and Future Prospects

Current main-detector analyses at the LHC span proper lifetimes $c\tau\sim0.1$ mm–100 m, with ATLAS and CMS setting $\mathrm{BR}(h\to LLP+LLP)\lesssim10^{-4}$–$10^{-2}$ for $c\tau$ in the cm–m regime (Jeanty et al., 22 Nov 2025, Collaboration, 30 Oct 2025, Gonski, 2022, Collaboration, 2024). Dedicated detectors (ANUBIS, MATHUSLA, FASER, DUNE ND) fill previously inaccessible windows up to $c\tau\sim10^7$ m, with sensitivity to branching ratios at $\mathcal{O}(10^{-6})$ and beyond. The highest reach is achieved for neutral, weakly-coupled, or low-mass LLPs, which evade prompt signatures; benchmark models include Higgs primordial decays, heavy neutral leptons (seesaw scenarios), hidden-valley glueballs, and low-scale SUSY NLSPs (Collaboration, 30 Oct 2025, Curtin et al., 2018, Qi et al., 17 Aug 2025).

Future improvements encompass advanced timing layers, high-luminosity upgrades, data-driven search programmability, and coordinated triggers extending coverage to soft or highly displaced signatures. Lower-energy experiments (beam-dumps, fixed-target) and astrophysical observations (BBN limits, CMB distortion, SN1987A cooling) complement the collider-based exploration of the full lifetime and mass parameter space (Jeanty et al., 22 Nov 2025).

7. Summary Table: Major LLP Detector Proposals and Sensitivity

Detector	Volume/Location	c$\tau$ Reach (m)	Sensitivity (BR$\,/\,\sigma$)	Target LLP Mass (GeV)
ANUBIS	$R=9.5$–$21$ m, ceiling above ATLAS	$0.1$–$10^{4}$	BR$(h\to XX)\sim10^{-6}$	$10$–$60$
MATHUSLA	$200\times200\times20$ m, surface	$100$–$10^7$	BR$(h\to XX)\sim10^{-5}$	$1$–$50$
FASER/FASER2	$L=1.5$–10 m, $R=0.1$–1 m, tunnel	$10^2$–$10^5$	Probe $\tan\beta$ up to $10^6$–$10^7$	$0.5$–$2$
DUNE ND-GAr	$L=5$ m, $D=5.2$ m, $\sim600$ m from source	$10$–$10^3$	BR(M$\to$LLP)BR(LLP$\to$vis) $\sim10^{-17}$	$0.1$–$2$

ANUBIS, MATHUSLA, FASER, CODEX-b, and DUNE define the experimental frontier for LLP searches, each tailored to specific mass, coupling, and lifetime regimes, but together spanning the full landscape of motivated BSM parameter space (Collaboration, 30 Oct 2025, Chou et al., 2016, Qi et al., 17 Aug 2025, Coloma et al., 2023). The LLP search program is thereby positioned as a central component of energy- and intensity-frontier experiments for the next decade and beyond.