SBND: Short-Baseline Near Detector

• SBND is a high-resolution liquid argon time projection chamber (LArTPC) detector that precisely measures unoscillated neutrino flux and interaction rates.
• Its advanced dual-drift design, detailed wire-plane assemblies, and integrated photon detection system deliver millimeter-scale spatial resolution and sub-percent energy accuracy.
• SBND plays a crucial role in precision oscillation measurements, neutrino interaction studies, and beyond-the-Standard-Model searches, guiding future experiments like DUNE.

The Short-Baseline Near Detector (SBND) is a high-resolution, high-statistics liquid argon time projection chamber (LArTPC) neutrino detector positioned 110 meters downstream of the Booster Neutrino Beam (BNB) target at Fermilab. As the near detector of the Short-Baseline Neutrino (SBN) program, SBND provides the crucial measurement of the unoscillated neutrino flux and interaction rates, enabling precision searches for eV-scale sterile neutrino oscillations, detailed ν–argon interaction studies, and a suite of beyond-the-Standard-Model (BSM) searches. SBND is also a primary technological demonstrator for future large-scale LArTPC experiments such as DUNE, operating with an active mass of 112 tonnes in a precisely engineered, dual-drift, membrane-style cryostat, and achieving millimeter-scale spatial and sub-percent energy resolution through a combination of advanced wire-plane construction, cold electronics, photon detection, and external cosmic tagging (Collaboration et al., 31 Mar 2025, Gao, 2019, Acciarri et al., 2020, Collaboration et al., 2024).

1. Detector Architecture and Instrumentation

SBND’s active volume measures 5.0 m (vertical) × 4.0 m (beam direction) × 4.0 m (drift), internally split by a –100 kV central cathode into two 2.0 m drift regions. Each drift volume features three Anode Plane Assemblies (APAs), each supporting three layers of 150 μm diameter Cu–Be wires tensioned to 7 N ± 1 N at 3 mm ± 0.5 mm pitch, and planarized to within ±0.5 mm over the 4 m × 2.5 m extent. The orientation of the wire planes (Y: vertical; U/V: ±60°) enables unambiguous 3D track reconstruction with ≤1 mm position uncertainty after deconvolution (Acciarri et al., 2020).

The wire readout is coupled to cold front-end electronics mounted directly on the APA frames, including 16-channel custom ASIC preamplifiers (programmable gain/shaper), 12-bit, 2 MS/s ADCs, and a low-power (5 W per 128-channel FEMB) digital multiplexing FPGA. Signals are delivered via 8 m cold copper twisted pairs to the warm interface for further processing, achieving an Equivalent Noise Charge (ENC) of 320–400 e⁻ (X/U/V planes) at 77 K and channel-to-channel RMS spread < 5% (Gao, 2019).

A dual photon detection system combines 120 eight-inch cryogenic PMTs—60 per drift region, primarily for triggering and t₀ extraction—and 192 X-ARAPUCA modules, which utilize dichroic filters and WLS bars for VUV light trapping and SiPM readout. TPB-coated reflective foils on the cathode, combined with patterned WLS coatings on the sensor faces, recover up to 400% more light at large drifts, delivering 15–24 PE/MeV (PMTs) and 8 PE/MeV (X-ARAPUCA), with uniformity to ≲20% (Collaboration et al., 2024). Timing resolution is determined as ≈2 ns for PMTs and 10–15 ns for X-ARAPUCAs across the relevant energy/position range.

A full-coverage Cosmic Ray Tagger (CRT) system constructed from dual-layer scintillator strip modules (10.8 cm × 2 cm × 2.3–4.1 m) surrounds the cryostat on five sides, providing 1–2 cm spatial and 1.8 ns temporal resolution, and >95% through-going muon tagging efficiency (Auger et al., 2016).

2. Cryogenics, High Voltage, Electric Field Calibration, and Quality Control

The LArTPC is hosted in a ProtoDUNE-style membrane cryostat with 0.45 m reinforced polyurethane foam insulation and a two-phase LN₂ heat shield, achieving a steady-state thermal balance at 89 K. Argon purity is maintained through gas and liquid recirculation; O₂-equivalent impurities are suppressed below 0.1 ppb to guarantee electron lifetimes τₑ > 2.5 ms, ensuring >90% charge survival over the maximal 2.0 m drift (Collaboration et al., 31 Mar 2025).

Field uniformity is sustained by a modular aluminum profile field cage and a 29-rung field shaping system (49.6 mm spacing). Field mapping and distortion correction are performed with a UV-laser calibration system: two precision-steered, 266 nm beams produce straight ionization tracks. Crossed-beam topologies at known intersection points enable direct measurement of 3D E-field distortions through local track displacement analysis, yielding spatial correction maps with ≤2 mm spatial bias and ΔE/E ∼1% uncertainty (Babu, 27 May 2025).

Mechanical and electrical tolerances in APA construction are strictly controlled: wire-pitch, tension, and flatness are verified at each production stage via laser tracking, photodiode tension measurement, electrical continuity/isolation tests (≤500 Ω wire resistance, ≥10 MΩ inter-wire isolation), and multiple cycles of cryogenic cycling (δTension ≤ 0.1 N, δR_cont ≤ 20 mΩ, δI_iso ≤ 700 nA). Such QA is essential to achieve induction-plane transparency >90%, position-dependent gain consistency, and thermal-mechanical robustness (Acciarri et al., 2020).

3. Beamline Configuration and Neutrino Flux Modeling

SBND samples the BNB at a baseline of 110 m, where the proton-beryllium target and horn-focusing system generate a beam with mean energy ≈0.8 GeV and >92% ν_μ content (6.9% $\bar\nu_\mu$, 0.6% ν_e + $\bar\nu_e$) at the peak (Collaboration et al., 31 Mar 2025). G4BNB, an updated GEANT4-based Monte Carlo framework, provides state-of-the-art flux predictions with full hadron production systematics (π ±, K ±, K^0, π^0, η) and neutrino ancestry tracking. The simulation achieves 8% higher peak flux and shrinks overall normalization uncertainties to ≈9–10% in the 0.2–1.5 GeV range, with dominant contributions from π^+ production (7.5%) and K^+ (4.0%) (Paton, 10 Jan 2025). Neutral meson (π^0, η) flux modeling enables sensitivity to exotic hidden-sector production channels.

SBND-PRISM extends basic flux measurements by exploiting the detector’s large front face to sample off-axis neutrino fluxes θ_OA = 0°–1.6°. This technique yields eight statistically independent “pseudo-monochromatic” fluxes, with covariance matrices publicly available, enabling cross-section and oscillation analyses largely independent of flux-shape and cross-section systematics (Abratenko et al., 27 Aug 2025).

4. Data Acquisition, Triggering, and Cosmic Background Rejection

The readout system, built around digitization at 2 MHz and 12-bit precision, employs a hierarchical FEMB/XMIT architecture. Continuous zero-suppressed “supernova” data and externally triggered “beam” data are processed in parallel streams, with FPGA-based real-time “trigger primitive” formation (per-plane charge-threshold crossing) and candidate clustering. Huffman compression, zero-suppression, and real-time region-of-interest (ROI) extraction achieve ×80 data reduction while retaining >99% of MIP-scale signals (Karagiorgi, 2019).

Self-triggering R&D allows for future TPC-only trigger operation, using primitive-based spatial/multiplicity and temporal coincidence. Michel-electron triggers have tested >95% efficiency and <1% false positive rates in analogous LArTPCs. Deep Neural Network augmentation for online event classification is implemented in DAQ farm prototype environments (Karagiorgi, 2019, Collaboration et al., 2020).

Cosmic backgrounds, dominated by O(10^4) muons per drift window, are suppressed by a combination of CRT tagging (timing-matched track extrapolation, >90% cosmic background rejection) (Auger et al., 2016), light-based T₀ matching, and advanced CNN-based semantic segmentation (UResNet, per-pixel cosmic/neutrino discrimination with mean IoU ≈0.8 and >90% cosmic rejection on test sets) (Collaboration et al., 2020). The CRT also provides calibration samples for dE/dx, drift velocity, and field uniformity diagnostics.

5. Physics Program: Oscillation Searches, Interaction Physics, and BSM Sensitivity

SBND anchors the SBN oscillation program through high-statistics measurement of the unoscillated νμ and ν_e spectra. In the three-detector fit (SBND, MicroBooNE, ICARUS), νμ→νe appearance and νμ disappearance are measured with sensitivity to sin^2(2θ_μe) > 2 × 10^−3 at Δm^2 ~ 1 eV^2 (90% CL after 10 × 10^20 POT) (Collaboration et al., 31 Mar 2025, Bass, 2017, Acciarri et al., 2015). Systematics from flux, cross-sections, and detector effects are strongly correlated and thus suppressed in the near-far comparison; the overall uncertainty in the oscillation measurement is driven to a few percent.

With ∼10^7 ν–Ar interactions/year, SBND enables per-channel ν_μ and ν_e CC/NC cross-section measurements with sub-percent statistical and <3% total systematic uncertainty in the 0.2–2.5 GeV band (Collaboration et al., 31 Mar 2025, Machado et al., 2019). Differential cross sections (e.g., dσ^CC/dQ^2) are extracted for CC0π, CC1π, NCπ^0, and rare topologies, facilitating precise tuning and validation of nuclear models for argon. The SBND-PRISM approach further mitigates model dependence by providing cross-section measurements over quasi-monochromatic, angle-dependent fluxes (Abratenko et al., 27 Aug 2025).

For BSM physics, SBND provides world-leading limits on lepton flavor violation (e.g., BR(π^+→μ^+ν_e) < 8.9 × 10^−4, BR(K^+→μ^+ν_e) < 3.2 × 10^−3) via the PRISM technique, and explores new parameter space for heavy neutral leptons, axion-like particles, and millicharged states, especially in planned off-target beam-dump modes post-2029 (Alves et al., 2024, Collaboration et al., 31 Mar 2025). Enhanced timing (σ_t ~ 1 ns), low detection thresholds (~2 MeV), and cosmic rejection support sensitivity to prompt and delayed exotic signatures.

6. Calibration, Reconstruction, and Performance Metrics

Comprehensive calibration is provided by UV-laser-based 3D E-field mapping, drift-velocity monitoring via external muons and in-situ laser tracks, and light- and charge-based energy-scale determination. Corrected data achieve vertex resolution improvement from ~5 mm (uncorrected) to ~2 mm, and calorimetric linearity restoration to within 5% of nominal across the TPC (Babu, 27 May 2025). Light-only 3D reconstruction using PMT and X-ARAPUCA signals achieves σ_X ~10–15 cm and σ_Z ~25 cm, supplementing TPC-based spatial information (Collaboration et al., 2024).

Trigger/DAQ performance has been validated to support >99% efficiency for beam neutrinos, with continuous “supernova” stream capability and zero-loss buffering. Timing synchronization (light, TPC, CRT) allows full exploitation of the BNB beam’s 19 ns spill structure, with full event timing resolution of ≈2 ns. The architecture is technology-parallel to DUNE DAQ/trigger concepts and supports DNN-based online data selection (Karagiorgi, 2019).

Systematic uncertainties on flux (<10%), cross-section (5%), and detector parameters (≈2–3%) are addressed via in-situ calibration, cross-detector comparison, and dedicated control samples. Comprehensive error budgets project ≤3% total normalization uncertainty per channel and <1% for relative unoscillated flux predictions (Collaboration et al., 31 Mar 2025, Acciarri et al., 2015).

7. Commissioning, Operations, and Future Directions

SBND began operations in July 2024, reaching stable beam data collection by December 2024, with ∼7,000 neutrino events/day and a target of ∼10^7 interactions by 2027. Post-2029 running plans include antineutrino and beam-dump configurations, aimed at both increasing the world’s $\bar\nu$–Ar cross-section sample and dramatically reducing neutrino-induced backgrounds for BSM searches (Collaboration et al., 31 Mar 2025).

As a key SBN and DUNE technology demonstrator, SBND validates full-scale deployment of modular wire plane assemblies, cold front-end electronics, X-ARAPUCA photon devices, UV-laser calibration, CRT integration, and scalable reconstruction frameworks (Pandora, Wire-Cell, DL-enhanced inference). Successful operational lessons are directly ported to DUNE’s planned far detector modules. Data, flux models (including G4BNB and PRISM covariances), and techniques are made public to support broader community engagement and cross-experiment analysis (Paton, 10 Jan 2025, Abratenko et al., 27 Aug 2025).

SBND’s integrated design and systematic control, together with unprecedented event statistics and mm-scale imaging, position it as the world-leading infrastructure for near-detector-based oscillation and neutrino interaction physics, as well as an essential stepping stone to the multi-kiloton LArTPC era.