X-ARAPUCA: Modular Photon Detector for DUNE

Updated 20 November 2025

X-ARAPUCA is a high-efficiency, modular photon detection system that enhances LArTPC performance through integrated wavelength shifters, dichroic filters, and SiPM arrays.
The design uses a combination of WLS slabs, reflective cavities, and optimized photon trapping to achieve significant PDE improvements and uniform light collection.
It is the baseline for DUNE’s far detector, enabling precise t₀ tagging and energy resolution, while also serving as a pathfinder for dark matter and other noble-liquid experiments.

The X-ARAPUCA (XA) is a high-efficiency, modular photon detection system originally designed for the Deep Underground Neutrino Experiment (DUNE) to enhance the sensitivity of liquid argon time projection chambers (LArTPCs) for both energy reconstruction and time-zero (t₀) tagging. By combining wavelength-shifting materials, dichroic filters, and highly reflective cavities with dense arrays of silicon photomultipliers (SiPMs), the X-ARAPUCA increases photon collection efficiency (PDE) beyond the capabilities of prior detector modules. It is now the worldwide baseline for DUNE’s far detector photon detection system and a technology pathfinder for other noble-liquid neutrino and dark matter experiments.

1. Device Architecture and Photon Collection Principle

The X-ARAPUCA builds on the ARAPUCA concept by introducing a wavelength-shifting (WLS) slab inside a highly reflective “photon trap.” This architecture systematically enhances photon collection via a sequence of wavelength conversion, optical filtering, and geometrically optimized trapping. The key elements are:

Entrance window: A glass or acrylic substrate coated on its outer surface with para-terphenyl (pTP, ~200–550 µg/cm²), serving as the first-stage WLS. It absorbs VUV (127 nm) LAr scintillation photons and re-emits at ~350 nm.
Dichroic filter: A multi-layer optical filter with a cut-on at ≈400 nm is deposited on the window’s inner side. It transmits photons below 400 nm (i.e., the 350 nm emission from pTP) but reflects longer wavelengths, preventing escape of WLS-reemitted >400 nm photons.
Internal WLS slab/bar: A light guide (commonly EJ-286PS or a specialized PMMA-BBT slab, thickness 2–5.5 mm, n≈1.5–1.6) doped with a secondary wavelength shifter (TPB or BBT) absorbs 350 nm light, re-emitting at ≈430 nm.
Photon guiding mechanisms: Photons at ≈430 nm are trapped by total internal reflection (TIR) within the slab or by repeated reflections off the dichroic filter and enhanced specular reflector (e.g., 3M Vikuiti/ESR, R>0.98) backplane, directing them toward the SiPMs.
SiPM arrays: SiPMs (e.g., Hamamatsu HPK HQR-75 or FBK Triple-Trench, cryogenically compatible, 6×6 mm²) are pressed edge-on to the slab, forming a continuous detection surface.
Optional reflecting sides: Non-instrumented edges are wrapped in Vikuiti to preserve TIR and minimize photon loss.

In baseline DUNE FD1 configurations, cells are ~80×100 mm² or up to 50×50 cm² in advanced concepts such as APEX (Marinho, 8 Mar 2025). Mechanical integration ensures robust optical coupling across cryogenic cycling.

2. Quantitative Photon Detection Performance

Absolute PDE is defined as the fraction of incident 127 nm photons on the entrance window resulting in photoelectrons at the SiPMs. The PDE factorization captures the chain of conversion and transport:

$\mathrm{PDE} = \varepsilon_{\mathrm{shift}}\cdot T_{\mathrm{filter}}\cdot P_{\mathrm{trap}}\cdot \varepsilon_{\mathrm{SiPM}}$

Where:

$\varepsilon_{\mathrm{shift}}$ is the pTP conversion efficiency (≥90–95%).
$T_{\mathrm{filter}}$ is the dichroic filter transmission at 350 nm (≈90%); removal of the filter can further raise throughput.
$P_{\mathrm{trap}}$ is the probability of a photon being retained by TIR or internal reflection from the 430 nm WLS emission (≈0.6–0.7 typical).
$\varepsilon_{\mathrm{SiPM}}$ is the SiPM PDE at 430 nm (≈45–60% at operating voltages, cryogenic temperatures).

Typical measured PDE values for DUNE SP and VD configurations, with SiPMs at recommended overvoltages, are summarized below (Álvarez-Garrote et al., 20 May 2024, Palomares et al., 2022, Botogoske et al., 15 Nov 2025, Corchado, 7 Feb 2025):

Configuration	PDE (%) @ 4.5 V OV	PDE Maximum (%)	Notes
DF-XA (with filter)	3.7 ± 0.3	4.7 ± 0.3 (@7 V OV)	Single-sided, standard slab
noDF-XA (no filter)	4.5 ± 0.4	5.8 ± 0.6 (@7 V OV)	~18% gain w/o dichroic
EJ-286PS WLS	1.8–2.4		Baseline slab, single-sided, FD1
PMMA-BBT/FB118 WLS	2.9 ± 0.1		+50% vs. EJ-286PS
G2P WLS bar	2.2–2.5		DUNE FD1 prototype
APEX tile	~2		50×50 cm², advanced FC wall coverage

Advanced R&D modules with larger tiles achieve position-dependent PDE uniformity to within 2–4% (Botogoske et al., 15 Nov 2025, Corchado, 7 Feb 2025).

3. Comparative Analysis and Design Optimization

Geant4 simulations, supported by laboratory measurements, confirm the relative gains of several key design parameters (Paulucci et al., 2019, Machado et al., 2018):

Internal WLS slab: Introduction of the WLS slab (X-ARAPUCA vs. ARAPUCA) yields a +30–50% collection efficiency increase due to TIR-guided photon transport, with fewer reflective bounces and reduced timing spread (Machado et al., 2018).
Filter removal: Eliminating the dichroic filter further increases absolute PDE by 10–20%. This is attributed to nonideal transmittance and angle-dependent cut-off behavior of real filters; the empirical PDE gain is ~18% as seen in vertical-drift designs (Botogoske et al., 15 Nov 2025, Corchado, 7 Feb 2025).
New WLS materials: PMMA-BBT (FB118) slabs exhibit quantum yields exceeding 90% and boost PDE by ≈50% over EJ-286, enabling DUNE FD1 to meet or exceed design light yield requirements for supernova sensitivity (Brizzolari et al., 2021, Souza, 2021).
Optical coupling: SiPM-bar gaps of >1 mm reduce PDE by up to 30%; optimal mechanical and optical contact is thus critical (Paulucci et al., 2019).
SiPM configuration: Coverage fraction, placement (long/short side), and type contribute 10–15% PDE variations; Hamamatsu devices slightly outperformed FBK in side-by-side benchmarks due to larger active area (Álvarez-Garrote et al., 20 May 2024).

Systematic uncertainties on PDE measurement are dominated by reference sensor calibration (±9–14%), geometric acceptance (±1.5–11%), SPE gain calibration (±2–10%), and cross-talk correction (±2–10%).

4. Applications in DUNE and Beyond

XA technology is the baseline for DUNE FD1 (horizontal drift), planned for ~6,000 units, and for the vertical-drift modules (~672–7,000 tiles depending on coverage) (Álvarez-Garrote et al., 20 May 2024, Corchado, 7 Feb 2025, Marinho, 8 Mar 2025). Its performance ensures:

Light yield: For DUNE FD1, 20 photoelectrons/MeV (PDE>1.3%) is satisfied for calorimetry and triggering; with FB118 or G2P slabs, typical yields are 22–28 PE/MeV at 4–5% PDE, supporting nucleon decay and SN burst sensitivity (Álvarez-Garrote et al., 20 May 2024, Botogoske et al., 15 Nov 2025).
Timing and t₀: Sub-ns to few‐ns timing spreads and reduced photon path length improve t₀ assignment, benefiting DUNE supernova burst and non-beam event analyses (Machado et al., 2018).
Energy resolution: For energies below 10 MeV, energy resolution from light (with 2.2–4.5% PDE) rivals that achievable with charge (TPC) readout, especially critical for SN neutrino physics (Souza et al., 2021).
Scalability: Modules retain stable efficiency (<1% drift over 10+ days, no evidence of bar or window degradation) and mechanical robustness across repeated cryogenic cycles. Modular assembly with cold electronics and fiber multiplexing (APEX) enables instrumenting large-volume detectors (Andreossi et al., 2023, Marinho, 8 Mar 2025).

XA modules are also under evaluation for DarkSide, LEGEND, SBND, and other experiments requiring high-efficiency LAr scintillation light readout (Brizzolari et al., 2021).

5. Design Variants and Performance Trends

Recent R&D has explored detailed XA variants to address the geometry and deployment constraints for both horizontal and vertical drift LArTPC modules:

Dichroic filter (DF) vs. noDF: Removal of dichroics provides robust PDE enhancement with minimal risk due to dominant TIR trapping. NoDF configurations now exceed baseline performance by ~18% (Botogoske et al., 15 Nov 2025, Corchado, 7 Feb 2025).
Single- vs. double-sided: Double-coated pTP windows provide <10% PDE gain compared to single-sided designs, implying limited justification for increased complexity (Botogoske et al., 15 Nov 2025, Corchado, 7 Feb 2025).
WLS bar optimization: Thicker (5.8 mm) bars with lower chromophore concentration do not significantly improve PDE but can flatten spatial response (good for large-area uniformity) (Botogoske et al., 15 Nov 2025).
Cold electronics and segmentation: Front-end cold transimpedance amplifiers and aggressive SiPM ganging maximize single-photoelectron sensitivity while reducing total channel count per detector volume (Brizzolari et al., 2021, Marinho, 8 Mar 2025).
Mechanical integration and scaling: The APEX design for DUNE FD3 targets 60% optical coverage per field-cage wall using XA tiles (0.25 m² each), yielding light yields of 109–180 PE/MeV and energy resolution down to 2–3% at 100 MeV (Marinho, 8 Mar 2025).

Table: Key PDE values (select DUNE FD configurations)

XA Version	PDE (%)	Notable Features
EJ-286, SS	1.8–2.2	Baseline slab, FD1
FB118, SS	2.9 ± 0.1	+50% vs baseline
G2P, SS	2.2–2.5	Custom slab, higher PDE
noDF-XA	4.5 ± 0.4	No dichroic, VD/FD
APEX tile	~2	50×50 cm², FC wall

6. Outlook, Limitations, and Future Developments

Ongoing directions for X-ARAPUCA optimization include:

Material R&D: Systematic fluor/concentration scans, custom filter development for sharper cut-offs, and alternative WLS compositions targeting improved quantum yield and mechanical resilience (Brizzolari et al., 2021, Botogoske et al., 15 Nov 2025).
Uniformity and systematics: Detailed mapping of PDE across full tile area; mechanical optimization of SiPM-bar alignment and pressure to remove ~10–20% position-dependent efficiencies (Álvarez-Garrote et al., 20 May 2024).
Cryogenic longevity: Extended-duration soaks and thermal cycling confirm no significant degradation of WLS or filter coatings; long-term stability is validated at ≤1% performance drift (Andreossi et al., 2023).
Integration with TPC readout: Advanced schemes (e.g., APEX) leverage power-over-fiber, high-density analog/digital multiplexing, and coverage >60% to further enhance both MeV and GeV-scale event reconstruction (Marinho, 8 Mar 2025).
Physics reach: Achieved PDE and coverage in the latest designs directly satisfy DUNE requirements for supernova, nucleon decay, and oscillation physics, with margin for TPC geometry constraints and device nonuniformities (Corchado, 7 Feb 2025, Botogoske et al., 15 Nov 2025).

The XA topology’s modularity, high PDE, and scalable integration with LArTPCs position it as the leading waveform-resolved photon collector technology for present and next-generation cryogenic neutrino and rare-event detectors.

References: