FORMOSA: Millicharged Particle Search at LHC

• Millicharged particles are hypothetical particles with a fractional electric charge (ε ≪ 1) predicted by models with hidden sector photons.
• FORMOSA employs a large, segmented plastic scintillator detector with precision timing and a four-fold coincidence requirement to reduce backgrounds.
• The experiment extends sensitivity in the forward region at CERN LHC, probing MCP masses up to ~45 GeV and charges as low as a few ×10⁻³.

A millicharged particle (MCP) is a hypothetical particle carrying an electric charge much smaller than the elementary charge e, typically parameterized as Q = εe with ε ≪ 1. The search for such particles has direct implications for physics beyond the Standard Model, including models with hidden sector photons/massless dark photons, extensions of quantum electrodynamics with charge quantization violation, and several classes of dark matter models. The FORMOSA (FORward MicrOcharge SeArch) experiment at the CERN LHC is designed to extend the sensitivity for MCPs by leveraging a dedicated, large-area scintillator detector placed in the far forward region of the LHC, where a highly collimated and intense flux of MCPs is produced. FORMOSA exploits precision timing, segmentation, and background rejection to probe unexplored regions of MCP charge–mass parameter space, building on advances in both detector technology and event modeling.

1. Theoretical Motivation and MCP Production at the LHC

Millicharged particles arise naturally in theories with a massless dark photon kinetically mixed with the Standard Model hypercharge field. In such scenarios, dark sector particles acquire a small but nonzero coupling to the photon, enabling their production in any Standard Model process that proceeds via an off-shell photon. At the LHC, MCPs are efficiently produced through Drell–Yan, proton–proton collisions, and vector meson decays, with production rates suppressed relative to ordinary leptons by a factor of (Q/e)². The relevant cross section can be written as

$$\sigma_{\text{MCP}} \approx \left(\frac{Q}{e}\right)^2 \cdot \sigma_{\text{SM}}$$

with additional mass-dependent corrections. For MCP masses up to tens of GeV, the high center-of-mass energy of the LHC ensures copious pair production across a broad pseudorapidity range. Notably, the forward region (pseudorapidity η ≈ 7) is especially advantageous: the anticipated MCP flux is enhanced by a factor of ∼250 relative to transverse locations (η ≈ 0) due to the near-uniform rapidity distribution of pair-produced MCPs (Foroughi-Abari et al., 2020, Citron et al., 17 Apr 2025).

In parallel, MCPs can also be produced in cosmic-ray air showers (via meson decay, proton bremsstrahlung, and Drell–Yan) (Argüelles et al., 2021, Wu et al., 3 Jun 2024), fixed-target beam dumps (Kelly et al., 2018), and rare decays of standard model mesons. These complementary production channels inform both the design and physics case for FORMOSA.

2. Scintillator-Based Detector Design and Signal Modeling

FORMOSA’s baseline design consists of a large, segmented plastic scintillator detector positioned ∼627 m downstream from the ATLAS interaction point at the Forward Physics Facility (FPF). The detector comprises:

• Four longitudinal layers, each a 20×20 array of bars (5 cm × 5 cm × 100 cm), cumulating to 1.9 m × 1.9 m × 4.8 m
• Each bar read out by a high-gain photomultiplier tube (e.g. Hamamatsu R7725), with single photoelectron (SPE) sensitivity
• Signal defined as the near-simultaneous (≤10–20 ns) coincidence of ≥1 SPE in each of the four layers along the track aligned towards the LHC interaction point

The detection principle exploits the fact that the energy loss of a MCP traversing the scintillator scales as dE/dx ∝ (Q/e)². For standard plastic scintillators (density 1 g/cm³, photon yield Yγ ≈ 1.1×10⁴ MeV⁻¹), the expected mean photoelectron (PE) yield per bar is

$$N_{\text{PE}} \approx \epsilon_{\text{det}} \rho_s L_s \langle -dE/dx \rangle Y_\gamma$$

where ε_det is the overall detection efficiency (∼10%). Sensitivity to fractional charges down to ε ∼ 10⁻³ is achievable due to the low detection threshold. For a 4-fold coincidence, the detection probability is (Foroughi-Abari et al., 2020, Citron et al., 17 Apr 2025):

$P_{\text{det}} = [1 - \exp(-N_{\text{PE}})]^4$

with backgrounds further suppressed by event-level and geometric cuts.

Upgrades leveraging high-performance scintillators (such as CeBr₃) have been demonstrated in the FORMOSA demonstrator (Citron et al., 17 Apr 2025), yielding up to ∼35x increased photon yield and faster response, enabling further reach for low ε.

3. Background Suppression and Demonstrator Results

FORMOSA’s projected background is controlled at an extremely low level, as required for rare event searches. Dominant backgrounds include:

• PMT dark noise (∼1 kHz per channel), suppressed via tight coincidences
• Cosmogenic muons, mitigated using hermetic veto panels and top/side/front layers
• Beam-induced afterpulsing, addressed with programmable deadtime windows following tagged beam-muon events

Monte Carlo (Pythia8, MadGraph5) and detailed Geant4 simulations, together with operational experience from the deployed UJ12 demonstrator, validate the event selection, DAQ stability, and robustness of the four-coincidence requirement. With all background sources considered and after efficient vetoing, the expected background rate is <1 event over the full projected HL-LHC data-taking period (~10 years) (Citron et al., 17 Apr 2025).

Crucially, the detector can be reconfigured in location or segmented for subdetector upgrades, with the demonstrated feasibility of integrating high-yield materials such as CeBr₃ in dedicated “high-sensitivity zones.”

4. Projected Sensitivity and Parametric Reach

For an integrated luminosity of 3000 fb⁻¹ (the HL-LHC target), the projected reach of FORMOSA (full-scale, FPF location) is expected to substantially exceed the anticipated performance of transverse detectors (e.g. milliQan). The enhanced forward flux, in combination with background rejection and effective area, allows FORMOSA to probe MCP masses up to ≈45 GeV and charges as low as ε ∼ few × 10⁻³ over a wide mass range (Foroughi-Abari et al., 2020, Citron et al., 17 Apr 2025).

The relevant sensitivity table (for illustrative values) is:

MCP mass (GeV)	Reach on ε (fractional charge)
0.1–1.0	few × 10⁻³
∼10	∼10⁻²
∼45	∼few × 10⁻²

For MCPs with even smaller ε (approaching ε ~ 10⁻⁴), the photon yield per traversal drops below unity, and new detection technologies or extended exposure are required.

The main sensitivity scaling is quadratic in the MCP charge: σ ∝ ε² (for production), with detection probability P_det scaling strongly for low PE yields, as above. Strict mass-dependent corrections apply for high-mass regions where production phase space is limited by the LHC kinematics.

5. Physics Impact, Complementarity, and Upgrades

FORMOSA’s experimental reach directly tests hidden sector extensions, quantization of electric charge (via direct limits on fractional charge), and diverse dark matter scenarios. Because production is proportional to ε² and the search is background-limited, FORMOSA is complementary to:

• Direct detection experiments (with higher MCP mass thresholds or less sensitivity to ε)
• Neutrino experiments exploiting atmospheric MCPs or beam dumps (Argüelles et al., 2021, Wu et al., 3 Jun 2024, Magill et al., 2018)
• Interferometric and optically levitated microsphere setups (probing ε down to ∼10⁻⁵ in bulk matter but not in high-flux, high-energy production regimes) (Moore et al., 2014, Gabrielli et al., 2016, Nugroho, 7 Jun 2024)

The ongoing demonstrator program at UJ12 establishes operational feasibility and provides a testbed for background studies, DAQ design, and validation of upgrade paths. Planned upgrades include hermetic side panels, high-yield scintillators for ultra-low ε, and data analysis enhancements, all designed to further suppress backgrounds and improve sensitivity (Citron et al., 17 Apr 2025).

6. Future Directions and Synergy

FORMOSA’s discovery potential can be further amplified by:

• Integration into the Forward Physics Facility (FPF), which affords a well-shielded, low-background environment, necessary infrastructure, and extended acceptance for additional forward-produced new physics signals (Foroughi-Abari et al., 2020)
• Upgraded scintillator technologies, DAQ bandwidth, and machine learning-based trigger strategies to optimize for extremely rare signal topologies
• Cross-calibration and joint analysis with LArTPC-based experiments (e.g. FLArE) at the FPF, which are optimized for MCP–electron scattering using different signal modalities (Kling et al., 2022)
• Exploring sensitivity to strongly interacting dark matter (SIDM) and other exotic signatures (e.g., via quadruple-coincidence topologies, or signatures suppressed by environmental attenuation in conventional detectors) (Foroughi-Abari et al., 2020)

A plausible implication is that the detection of MCPs at the sensitivity enabled by FORMOSA would set novel constraints, or discover new physics, in sectors linked to charge quantization, hidden photons, and the detailed structure of the dark matter sector. The combination of a robust design, operational validation, and forward geometry positions FORMOSA as a leading experiment in the worldwide search for subelectronic charged particles.