Response of wavelength-shifting and scintillating-wavelength-shifting fibers to ionizing radiation

We report results of characterizing the response and light transport of wavelength-shifting (WLS) and scintillating-wavelength-shifting (Sci-WLS) fibers under irradiation by radioactive $α$, $β$, and $γ$ sources. Light yield and light transmission were measured for the WLS fiber BCF-91A from Saint-Gobain and for a new Sci-WLS fiber EJ-160 from Eljen Technology. The two variants with different fluor mixtures, EJ-160I and EJ-160II, exhibited approximately five and seven times higher light yield than BCF-91A, respectively, while their attenuation lengths were 3.80,m for BCF-91A, 4.00,m for EJ-160I, and 2.50,m for EJ-160II.

💡 Research Summary

The paper presents a systematic study of the optical response of conventional wavelength‑shifting (WLS) fibers and newly developed scintillating‑wavelength‑shifting (Sci‑WLS) fibers when exposed to ionizing radiation from α, β, and γ sources. The motivation stems from the widespread use of plastic WLS fibers in large‑scale particle‑physics experiments (e.g., MINOS, NOvA, T2K, LHCb) and the emerging requirements of next‑generation projects such as LEGEND‑1000 and ePIC, where fibers must not only collect and transport light efficiently but also meet stringent radiopurity standards and potentially serve as active radiation detectors.

Three fiber types were examined: the established BCF‑91A WLS fiber from Saint‑Gobain, and two variants of a new Sci‑WLS fiber from Eljen Technology, designated EJ‑160I and EJ‑160II. All fibers have a 1 mm × 1 mm square cross‑section, a polystyrene core, and a PMMA cladding (0.03 mm for BCF‑91A, 0.04 mm for the EJ‑160 series). Their absorption peaks lie at 424 nm (BCF‑91A) and 427 nm (EJ‑160), while emission peaks are at 494 nm and 490 nm, respectively, providing comparable spectral matching to the SiPM used.

The experimental setup employed 1.4 m long fiber samples with both ends optically coupled to Hamamatsu S13360‑3050CS silicon photomultipliers (SiPMs) using BC‑630 optical grease. The SiPM bias and readout were handled by a custom board with a transimpedance amplifier; waveforms were digitized on a LeCroy WaveRunner HR 660Zi oscilloscope. Three radioactive sources were used: a 90Sr β source (4.8 µCi), a 22Na γ source (3.4 µCi, 511 keV line), and a 241Am α source (1.0 µCi, 5.486 MeV). For β and γ measurements the source was moved along the fiber at 13 positions ranging from 5 cm to 133 cm, allowing a detailed mapping of light yield versus distance. For α measurements, due to the short range of alphas, fibers of varying lengths (5 cm to 138 cm) were irradiated from one side while only one SiPM read out the opposite end.

The number of photoelectrons (p.e.) recorded by the SiPMs was extracted from the pulse‑height distributions. To describe the distance dependence, the data were fitted with a double‑exponential model:
I(x) = I_long exp(‑x/λ_long) + I_short exp(‑x/λ_short),
where λ_long corresponds to core‑guided light (long‑range attenuation) and λ_short to cladding‑guided light (short‑range attenuation). This functional form captures the well‑known two‑component propagation in plastic fibers: a dominant early component from the higher‑trapping‑efficiency cladding mode that decays within the first meter, and a slower, lower‑amplitude core mode that persists over several meters.

Key results are summarized in Table 2 of the paper. For β irradiation, the extrapolated light yields at zero distance are 12.7 p.e. (BCF‑91A), 64.0 p.e. (EJ‑160I), and 87.7 p.e. (EJ‑160II). For γ irradiation the yields are 10.0 p.e., 55.7 p.e., and 76.9 p.e., respectively. For α irradiation the yields are 28.5 p.e., 81.6 p.e., and 103 p.e. Thus EJ‑160I and EJ‑160II provide roughly a factor of five and seven higher photoelectron counts than the standard BCF‑91A for β particles, and similar enhancements for γ rays. The α‑induced enhancement is smaller (≈3×) because high‑dE/dx alphas cause stronger quenching in the polystyrene‑based scintillators.

Attenuation lengths derived from the double‑exponential fits (or taken from a previous 3 m‑long fiber study for λ_long) are: λ_long = 3.80 m (BCF‑91A), 4.00 m (EJ‑160I), and 2.50 m (EJ‑160II); λ_short values are all around 0.07–0.12 m. The longer λ_long of EJ‑160I makes it attractive for applications requiring long‑range light transport, whereas EJ‑160II, despite its shorter λ_long, delivers the highest initial light yield, which can be advantageous for relatively short detector geometries such as the 1.4 m fiber length used in LEGEND‑1000.

Systematic uncertainties were evaluated by fully rebuilding the setup ten times, including re‑polishing fiber ends, re‑applying grease, and repositioning sources. Statistical uncertainties are negligible because each measurement accumulated tens of thousands of events.

The authors discuss the implications for low‑background experiments. The high light yield of the Sci‑WLS fibers could enable fibers to serve as active vetoes or self‑tagging elements for intrinsic radio‑impurities, potentially relaxing material‑purity constraints. The observed intrinsic scintillation in the aromatic polymer matrix (polystyrene) explains why even a pure WLS fiber (BCF‑91A) produces a measurable signal under ionizing radiation.

In the conclusions, the paper emphasizes that the EJ‑160 series outperforms the benchmark BCF‑91A in terms of photoelectron yield while offering comparable or acceptable attenuation lengths. Ongoing work includes extending attenuation measurements to longer fibers, expanding the comparative study to other commercial WLS fibers (e.g., Kuraray Y‑11), refining the fluor formulation to improve radiopurity, and developing a comprehensive Monte‑Carlo framework for light transport that has already been validated against Kuraray fibers. These efforts aim to deliver optimized Sci‑WLS fibers tailored for the stringent performance and background requirements of next‑generation rare‑event searches.