Commissioning of a mobile neutron spectrometer for LNGS

Environmental neutrons are a source of background for rare event searches in underground laboratories. Since the majority of the neutron background comes from the cavern walls due to the intrinsic radioactivity of concrete and rock, the flux is known to be time and location dependent. Therefore, a precise knowledge of the spectrum and of the total flux is needed to devise shielding and veto mechanisms for rare event searches. Here ALMOND (An LNGS Mobile Neutron Detector) is presented. It is a mobile neutron spectrometer, based on capture-gated spectroscopy and comprised of an array of plastic scintillator bars wrapped with gadolinium foils. The detector has been calibrated with Americium-Beryllium source at Karlsruhe Institute of Technology and with an Americium-Boron source and a D-D generator at ENEA Frascati. The results of the neutron calibration with the time of flight method and the D-D generator are shown here, alongside the first results on capture time profile. Moreover, the first results from the neutron background run in Hall A at LNGS are presented.

💡 Research Summary

The paper presents the design, calibration, and first underground deployment of ALMOND (An LNGS Mobile Neutron Detector), a portable neutron spectrometer intended for precise background neutron measurements at the Laboratori Nazionali del Gran Sasso (LNGS). The authors motivate the work by emphasizing that underground rare‑event searches (dark matter, neutrinoless double beta decay, etc.) are limited by neutrons produced in the surrounding rock and concrete via spontaneous fission of ^238U and (α,n) reactions on light nuclei. Because the concentration of radioactive elements and the water content of the cavern vary with location and season, the neutron flux is both spatially and temporally non‑uniform, demanding an instrument that can map the spectrum at many points with a consistent methodology.

ALMOND’s hardware consists of 36 EJ‑200 plastic scintillator bars (5 × 5 × 25 cm³) each wrapped in a 100 µm gadolinium foil and coupled to a low‑background 3‑inch ET 9302B photomultiplier tube. Gadolinium captures thermal neutrons with a large cross‑section and releases an ∼8 MeV gamma cascade, which is well above the natural gamma background (≤2.6 MeV) and therefore provides a clean capture tag. The array is surrounded by a 16 mm lead shield to suppress external gammas that could otherwise be misidentified as capture events. The data‑acquisition system triggers on a low‑energy proton recoil (≥20 keV electron‑equivalent) and then records all waveforms within a ±100 µs window at 62.5 MS/s. A secondary high‑energy threshold (3 MeV) is applied offline to isolate the gadolinium capture signal.

Calibration was performed in three stages. First, at Karlsruhe Institute of Technology (KIT) an AmBe source was used together with a BGO gamma detector to implement a time‑of‑flight (ToF) measurement. The simultaneous emission of a 4.4 MeV gamma and a neutron allowed tagging of the neutron arrival time, from which the scintillator light yield versus neutron kinetic energy was extracted. The data were fitted with Birks’ law, yielding a Birks constant consistent with earlier studies. After the full 36‑module array was assembled, a second AmBe run was carried out without ToF tagging; instead, the capture‑time distribution was obtained by selecting events with a single pulse cluster (minimum 5 samples ≈90 ns separation) and subtracting accidental coincidences.

A second calibration campaign took place at the Frascati Neutron Generator (FNG). Here, a deuterium‑deuterium (D‑D) neutron generator provided quasi‑monoenergetic neutrons (nominal 2.4 MeV, effectively 2.2 MeV at the detector angle >150°). The ToF analysis showed that recoil signals extend up to ~700 keVee, beyond which the background follows an empirical power law (∝E⁻¹·²). By differentiating the background‑subtracted spectrum, the proton‑recoil edge was identified at 654 ± 4 keVee, confirming the energy calibration. An AmB source was also used, and the capture‑time profile obtained matched that from the earlier AmBe calibration, demonstrating the stability of the capture‑gate timing.

Following calibration, ALMOND was deployed underground in Hall A of LNGS in February 2025. Data were collected with a 40 µs capture window, a 3 MeV capture threshold, and a 20 keVee recoil threshold. The authors excluded a region contaminated by a nearby neutron source (highlighted in yellow in their figure) and reported daily rates of neutron candidates and accidental coincidences. The measured ambient neutron flux was on the order of 10⁻⁶ n cm⁻² s⁻¹, comfortably within the design specification (10⁻⁶–10⁻⁵ n cm⁻² s⁻¹). A subsequent run began in Hall C in August 2025, and plans are outlined to extend measurements to additional LNGS halls, thereby building a comprehensive spatial map of the neutron background.

In conclusion, ALMOND demonstrates that a mobile, modular, capture‑gated neutron spectrometer can deliver consistent, high‑resolution neutron spectra across different locations in a deep underground laboratory. Its ability to separate prompt recoil signals from delayed gadolinium capture gammas, combined with robust calibration using both AmBe/AmB sources and a D‑D generator, provides the precise spectral information required for optimizing shielding and veto systems in rare‑event experiments. Future work will focus on automation, real‑time data transmission, and coordinated multi‑site deployments to further refine background models and support the next generation of low‑background physics searches.