Skip to content

Diamond Detector Development for Neutron Measurements at the n TOF Facility (CERN)

FieldValue
JournalHNPS Advances in Nuclear Physics
Published2026-02-24
AuthorsK. Kaperoni, Μαρία Κ. Διακάκη, Michael Bacak, Christina Weiss, Erich Griesmayer
DOI10.12681/hnpsanp.8965

A 50 µm single-crystal diamond detector measured the ⁶Li(n,t)⁴He reaction at CERN’s NEAR station, less than 3 m from the spallation target. At the NEAR collimator exit, the neutron flux reaches about 1.5 × 10⁹ n/cm² per standard pulse of 7 × 10¹² protons. Low-intensity pulses resolved higher neutron energies after dedicated and parasitic pulses saturated above about 2 × 10³ ns time of flight.

NEAR receives a mixed field from 20 GeV protons striking a thick lead spallation target. The white neutron spectrum spans more than eleven orders of magnitude in energy. EAR1 and EAR2 provide flight paths of approximately 185 m and 20 m, respectively. At the collimator exit, the NEAR flux is approximately two orders of magnitude above EAR2. NEAR’s short path limits neutron energy resolution and exposes active detectors to severe radiation. NEAR was commissioned in 2021 for measurements close to the target.

Most NEAR experiments use activation samples for astrophysical cross sections or material irradiation for engineering studies. Active detectors face a harder test because they operate in the mixed neutron and gamma field. The conversion layer was 1.8 µm ⁶LiF with 95% ⁶Li enrichment. It generated tritons and alpha particles through the ⁶Li(n,t)⁴He reaction after neutron absorption.

The authors irradiated 50 µm, 140 µm, and 500 µm single-crystal CVD detectors with a ¹³⁷Cs source emitting 661.6 keV gamma rays. Measured interaction probability and deposited charge increased with detector thickness. The 500 µm sensor showed a Compton edge at 0.48 MeV. The 50 µm and 140 µm spectra did not show that edge.

The measurements linked thicker diamond to greater gamma contamination during neutron measurements. The 50 µm sensor was selected for later tests because it limited gamma contributions. A cited Geant4 study for fast neutrons also found 50 µm balanced neutron sensitivity and gamma suppression. That fast-neutron result came from simulation, while the source spectra were experimental. The gamma exposure was needed because correct discrimination thresholds depend on the gamma background. The gamma test preceded neutron-beam exposure and supplied direct thickness evidence.

The selected 50 µm detector was tested at the TRIGA reactor in Vienna. It stood at the exit of the reactor beam line. It operated at -50 V, corresponding to an average electric field of 1 V/µm. The Geant4 model used a 25.3 eV monoenergetic beam with a flux of approximately 1.5 × 10⁴ n/cm²/s. The thermal field was well characterized.

The measured spectrum showed a distinct triton peak at approximately 2.5 MeV. Alpha signals extended to about 1.3 MeV. Both features agreed with the simulated spectrum, which validated the modeled geometry and material configuration. The simulated geometrical triton detection efficiency was 4.45%. The paper states a statistical error of 9 × 10⁻⁵ for that efficiency. The thermal test established charged-particle detection before white-beam exposure.

EAR2 offered a well-characterized neutron flux and a 20 m vertical flight path. The well-characterized flux made EAR2 suitable for the first in-beam test of the detector and amplifier. The detector sat on the beam axis inside aluminum shielding. A 1 m cable fed a 2 GHz spectroscopic amplifier with 40 dB gain. The amplifier had been developed for the high rates expected at NEAR. The detector operated at -50 V, and the amplifier output entered the n_TOF data acquisition system.

Signal amplitude was recorded against time of flight in two-dimensional histograms. Dedicated pulses delivered 8.5 × 10¹² protons per pulse. Parasitic pulses delivered 4.5 × 10¹² protons per pulse. The triton peak remained between 80 and 120 mV for both conditions. Alpha signals reached about 50 mV. At a time of flight near 2 × 10³ ns, the ⁶Li resonance appeared at roughly 200 keV neutron energy. The 20 m path separated tritons and alphas clearly at higher times of flight, corresponding to lower neutron energies. The authors treated EAR2 as a preparatory step for the main NEAR experiment.

The 70.8 m cable introduced measurable signal degradation. A 2 ns pulser waveform was generated at the DAQ, sent to NEAR, and reflected back. The waveform lost 41.5% of its amplitude. Its FWHM increased by a factor of 2.5. Measured triton signals also broadened by approximately 2.8 relative to EAR2. The comparison used a 20 m cable at EAR2 and a 70.8 m cable at NEAR. These results quantified the signal-quality cost of the NEAR cable length.

At NEAR, the detector was 12 cm from the marble shielding wall and 32 cm from the collimator exit. It was aligned parallel to the neutron beam axis. The detector retained the -50 V bias and the same 2 GHz, 40 dB amplifier used at EAR2. The amplifier connected directly to the detector. Its output crossed the 70.8 m cable to the DAQ.

The NEAR histograms used signal amplitude and time of flight as axes. Dedicated and parasitic pulses showed significant saturation up to about 2 × 10³ ns time of flight. This interval corresponds to neutron energies above 8 MeV. The low-intensity condition used approximately 10¹⁰ protons per pulse. Under this condition, the ⁶Li resonance became visible at higher neutron energies. The authors attributed the recovery to lower gamma flash and neutron pile-up. The triton amplitude was lower than at EAR2, mainly because of attenuation along the 70.8 m cable. The broader energy distribution reflected the poorer time-of-flight resolution at the short NEAR path. The result demonstrated the first NEAR neutron-beam measurement through the ⁶Li(n,t)⁴He reaction.

Engineering implications and remaining work

Section titled “Engineering implications and remaining work”

The ¹³⁷Cs tests support 50 µm single-crystal diamond when gamma background limits discrimination. Thickness selection reduced gamma interactions, but it did not remove high-rate saturation. The 70.8 m cable also reduced pulse height and broadened the signal before data acquisition. The tested NEAR configuration kept the detector and amplifier directly connected. Low-intensity pulses extended usable neutron-energy coverage when dedicated and parasitic pulses saturated.

The thermal result provides a geometry reference, but its 4.45% efficiency came from simulation rather than an in-beam calibration. No NEAR-specific efficiency calibration or final flux uncertainty is reported. The result covers in-beam operation and reaction discrimination. Flux and beam profile characterization remained future work. The 3 m flight path will continue to limit neutron energy resolution. The data expose cable attenuation and high-rate saturation as constraints. The detector is intended to characterize the NEAR station’s neutron flux and beam profile.

The device needs a 50 µm single-crystal diamond sensing element paired with a 1.8 µm, 95% ⁶Li-enriched ⁶LiF converter for the ⁶Li(n,t)⁴He measurement. 6CCVD’s diamond materials catalog lists single-crystal diamond grown by MPCVD with thicknesses from 0.1 to 500 µm, covering the sensor thickness used here. The engineering support page offers custom fabrication support for the detector geometry and assembly.

View Original Abstract

Diamond is a highly promising material for neutron detection due to its radiation hardness, fast response, and high signal-to-noise ratio. A single-crystal diamond (sCVD) detector with a 6LiF conversion layer was developed in collaboration with CIVIDEC Instrumentation for in-beam measurements at the n_TOF facility at CERN. The effect of the diamond detector thickness on the response to 𝛾-rays was studied with a 137Cs source. Additionally, measurements with thermal neutrons at the TRIGA reactor (Atominstitut Vienna) confirmed the proper detection of the recoils of the 6Li(n,t)4He reaction, by the developed detection system. In-beam tests at the EAR2 experimental area demonstrated stable performance under a white neutron beam, while the final easurement at the NEAR station validated the detector’s suitability for neutron flux characterization in challenging high-radiation environments.