Online program Overview Session: NSS-01

Abstract

Background γ-ray fields complicate neutron measurements for Nondestructive Assay (NDA) of Nuclear Materials and the next generation of advanced reactor designs, including non-light water (non-LWR) and Small Modular Reactors (SMRs). Neutron sensitive proportional counter design for high γ-ray fields often focus on optimizing ion collection time and integrating low-Z materials. In these counters, the metal cathode produces most of the Compton scattered electrons, which deposit unwanted energy and signal in the gas. The long path length for Compton electrons in the counter gas means the counter gas pressures may be optimized to increase the ratio of ion-to-electron energy deposition. Here we present gas mixture and pressure optimization for ion-to-electron energy deposition using SRIM simulations. The simulation output is then tested experimentally for its effect on neutron sensitivity in γ-ray fields between 1 R/Hr and 100 R/Hr. The ion-to-electron energy deposition ratio shows promise as an additional design consideration for neutron sensitive proportional counters in γ-ray fields, even though the experimentally measured optimum pressure of 0.5 atm for He/N₂ 95/5 is lower than the simulated optimum pressure. Future experiments with a current sensitive preamplifier (CAEN A1427) will improve neutron counting for 1 to 100 R/Hr γ-ray fields.

Abstract

High-pressure ionization chambers (HPICs) are widely used for radiation monitoring owing to their high sensitivity and mechanical robustness. However, under low-dose conditions, the inherently small signal results in a low signal-to-noise ratio, making accurate characterization of detector response challenging. Therefore, detailed analysis of radiation interaction and charge transport processes in the detector is required for accurate characterization and optimization of HPIC performance. In this study, a simulation framework was developed to investigate and optimize the signal response characteristics of HPICs. A coupled simulation framework integrating Geant4 and Garfield++ is used to model the full signal formation process, including radiation dose deposition and charge transport under applied electric fields. A Kyocera HPIC was used to measure the current response as a function of applied voltage under irradiation from a 95 μCi Cs-137 source. In the Geant4 simulation, the HPIC geometry was modeled to calculate the spatial distribution of dose deposition within the gas volume, which is subsequently converted into ion pair distributions using the W-value. The electric field is modeled in Garfield++ to simulate charge transport under applied voltages, enabling the calculation of the resulting current response. The collected charge in the anode shows a plateau behavior at approximately 64.2 pA, while the experimental results exhibit a saturation around 54.8 pA. It is predicted that the discrepancy is attributed to additional physical effects such as ion recombination, feedthrough effects, and leakage currents. These findings demonstrate that the proposed approach provides a reliable framework for accurate analysis and optimization of HPIC.

Abstract

A worldwide Lepton Collider Time Projection Chamber community of the physicists was working to realize an exceptional physics program of the energy-frontier, electron-positron collision, especially for Future Circular Collider (FCC-ee), Circular Electron Positron Collider (CEPC) and the International Linear Collier (ILC). In all proposed detector concepts, the Time Projection Chamber (TPC) serves as the main tracking detector, complemented by the silicon tracking detectors both inside and outside the TPC, all embedded within a 3 T solenoid field. Many extensive R&D program for the track detector has been developed within the collaboration of LCTPC.

TPC performance is satisfactory in Higgs mode, but the high luminosity Z pole running poses substantial technical challenges. Based on MDI background simulation data to study ion-induced space charge effects, we have conducted a detailed assessment of TPC technical challenges and potential solutions, particularly under the high luminosity Z pole running conditions. In this talk, the feasibility and update progress of high precision TPC as the main track detector for e+ e− collider will be presented. TPC detection aims the good separation power with the cluster counting to be considered and the simulation results of the pad/pixelated TPC technology will be given. Compared with the pad readout using the simulation, the high granularity readout TPC options will obtain the better spatial resolution of single electrons, the very high detection efficiency in excellent tracking and good PID performance (less than 3σ).  A large prototype TPC in a 1.0T magnetic field, which allow to accommodate up to seven identical Micro Pattern Gaseous Detector (MPGD) readout modules of the TPC being studied using the 5GeV electron beam at DESY, we will present the track reconstruction performance results and summarize the next steps of the pad/pixelated TPC technology for the future lepton colliders.

Abstract

This paper presents a semi-analytical method for sensitivity estimation of multilayer parallel-plate boron-coated ionization chambers. The model combines: (1) an overlap-geometry calculation for the distribution of crossed boron layers under oblique incidence, and (2) a microscopic-simulation-based depth-response function for single-layer effective signal generation. By coupling geometric crossing statistics, boron attenuation, and depth response, detector sensitivity can be rapidly evaluated versus incidence angle, boron thickness, and plate spacing. Compared with Geant4, the method reproduces both trend and magnitude of angular sensitivity, with better agreement at large angles. The framework is suitable for fast parameter screening before high-fidelity simulation.

Abstract

Pulse shape discrimination (PSD) is an effective technique for identifying the interacting particle (e.g., gamma rays or neutrons) in scintillation detectors.
In the simplest cases (e.g., the Cs₂LiYCl₆ scintillator, showing a large difference between gamma and neutron signals), the PSD procedure can be performed using the "charge integration" (CI) method, based on rectangular weighting functions. However, for scintillators with smaller differences in signal shape (e.g., the LaCl₃:Ce scintillator), different methods must be used.
Moreover, when digital acquisition systems are used, sampling the photodetector’s analog signal at a finite time step can significantly affect the PSD results. Indeed, apart from the effects of electronic and quantization noise, the main cause of overlapping of the PSD regions of interest (ROI) of fast scintillator signals processed in the time domain is the temporal variability (ΔT) of the sampling comb relative to the onset of the analog pulse. In the case of relatively high DAQ's acquisition frequency (e.g., 500 MHz), increasing it further to mitigate this issue is not practical, and alternative techniques must be used.
One solution is to proceed via the absolute value of the discrete Fourier transform (DFT) of the signal, which completely solves the temporal variability issue, though at the cost of significant computational burden.
This paper presents a new analysis technique that solves the problem with a negligible computational overhead, making it especially suitable for online applications, by just linking the information on the time shift ΔT to the detector signal, and linearly filtering the whole dataset.
The filter weighting function is determined with the DPLMS optimal filter synthesis algorithm, employed here for the first time for this purpose.
The PSD results obtained with this technique, suitable as well for energy and time estimations, are here shown for experimentally acquired  LaCl₃:Ce scintillator signals.

Abstract

Near infrared (NIR) emitting scintillators have attracted attentions recently for radiation monitoring at high dose fields. In this work, NIR scintillation properties of Yb3+ and Nd3+ doped fluorapatite (Ca10(PO4)6F2, FAP) transparent ceramics with different doping concentration were investigated. These materials have been originally developed for laser applications. The most famous laser material, Nd:YAG transparent ceramic, can show good scintillation performance both UV-VIS emission and NIR emission, and following such a study, good performance can be expected in FAP transparent ceramic. When X-rays were irradiated, they exhibited 1000 nm (Yb) and 1064 (band) emission bands, and their scintillation decay time resulted several micro hundreds s. The most important property for high dose field monitoring, the lower detection limit (LDL), was investigated by coupling with InGaAs photodiode. Yb and Nd-doped FAP showed 1.2 and 0.6 mGy/hr, respectively. In the conference, scintillation properties as well as photoluminescence properties will be presented, and compared with other NIR emitting scintillators.

Abstract

Halide perovskites are promising scintillator materials for radiation detection and medical imaging because of their excellent optical and scintillation properties. Since the surface quality of single-crystal scintillators strongly affects light extraction and detector performance, polishing is an essential step in scintillator fabrication. However, conventional wet polishing processes are not suitable for water-sensitive crystals such as Cs₃Cu₂I₅ (CCI). In this work, hydrophobic solvents were investigated as polishing media for water-sensitive CCI single crystals. Three hydrophobic solvents—squalane, silicone oil, and mineral oil— were used as dispersion media for 1 μm alumina abrasives, followed by toluene rinsing after polishing. Among them, squalane showed the best polishing performance, yielding the lowest surface roughness (R_q=17.29 nm), the highest transmittance of 78.21%, and the narrowest rocking-curve FWHM of 0.025°. In addition, sequential polishing in squalane using P2000 sandpaper, P4000 sandpaper, and 1 μm alumina enhanced the light yield and improved the energy resolution from 6.53% to 5.83%. These results identify squalane as a promising polishing medium for water-sensitive halide perovskite scintillators.

Abstract

Time-of-flight positron emission tomography (ToF-PET) has boosted PET image quality significantly in the last years. During its first inception it was built with cross-luminescent BaF₂ crystals, which have since been replaced by LYSO. Recently, cross-luminescence has regained popularity with new scintillators being developed. However, BaF₂ and the recently developed Cs₂ZnCl₄ could not improve on the LYSO state-of-the-art coincidence time resolution (CTR) (56 ps for 2x2x3 mm³ and 95 ps for 3x3x19 mm³). As Cs₂ZnCl₄ shows a significant Cherenkov production, we combined it with the DIGILOG approach to boost its timing performance. The DIGILOG approach segments the sensor into μSiPMs and has shown an improved timing performance for BGO due to its Cherenkov radiation and an improved SPTR. Here, we simulated a 3x3x3 mm³ and a 3x3x20 mm³ Cs₂ZnCl₄ crystal coupled to seven different μSiPM arrays. The simulation included noise factors like dark counts, cross-talk, SPAD dead time and analog noise. It was designed with a PDE of 60 % and an SPTR of 25 ps. This resulted in an improved CTR of 34+/-4 ps for an array of 20x20 μSiPMs coupled to a 3x3x3 mm³ Cs₂ZnCl₄ crystal compared to the non-segmented SiPM at 48+/-3 ps. Furthermore, for the 3x3x20 mm³ crystal, a CTR of 64+/-4 ps was reached with a 10x10 μSiPM array, while the non-segmented SiPM showed a CTR of 84+/-3 ps. The results of the segmented SiPMs improve on the LYSO state-of-the-art. Additionally, they demonstrate that the DIGILOG SiPM approach improves timing performance not just for BGO, but also for other scintillators with high Cherenkov yield, ultra-fast emission and low light yield. We also provide motivation to optimize sensors for Cs₂ZnCl₄, which could enable better timing performance than the LYSO state-of-the-art.

Abstract

The temperature dependence of pulse shape discrimination performance was investigated for three commercial silicon photomultiplier technologies coupled to an EJ-276D organic scintillator. Measurements were conducted at 22 °C, 30 °C, 40 °C, and 50 °C using PuBe sources. Pulse shape discrimination was quantified using the tail-to-total ratio and figure of merit. For all devices, the figure of merit increased at 30 °C. The Broadcom device exhibited the highest performance, increasing from 1.12 at 22 °C to a maximum of 1.17 at 30 °C before decreasing to 0.76 at 50 °C. The Onsemi and Hamamatsu devices showed smaller improvements, with figure of merit values increasing from 0.86 to 0.89 and 0.84 to 0.90, respectively, between 22 °C and 30 °C. At 50 °C, all devices exhibited reduced performance due to increased silicon photomultiplier thermal noise. Temperature-dependent changes in pulse shape were observed, with increasing temperature producing faster decay and modifying the balance of prompt and delayed scintillation light, improving neutron-gamma separation at intermediate temperatures. Despite degradation at higher temperature, all devices maintained effective discrimination up to 50 °C. These results demonstrate that silicon photomultiplier-based detectors are capable of robust neutron-gamma discrimination at elevated temperatures relevant to safeguards applications.

Abstract

Cerium-doped gadolinium gallium aluminium garnet (GAGG:Ce) is a novel scintillator with excellent scintillation properties. It is suitable for use in various basic science research and applications. Because the scintillation light is quenched for charged particle detection, it is necessary to use beams from accelerators to calibrate the energy response. Measurements of energetic charged particles incident on three GAGG:Ce crystals were performed using accelerated protons, ⁴He, and ¹²C at 0^oC and 20^oC in consideration for detecting charged particles in the earth's radiation belts for space weather modeling and for deploying in dosimetry devices in ion therapy, respectively. The Birks factor which quantitatively describes the extent of quenching was deduced from the experimental data. For all three GAGG:Ce scintillators, the Birks factor decreases with the atomic number Z of the incident ions. Furthermore, less light quenching, a smaller Birks factor, was observed for GAGG:Ce at 0^oC. The quenching for the Mg-codoped GAGG:Ce which has a shorter decay time constant, was found to be less temperature dependent compared to the other two GAGG:Ce crystals.

Abstract

Metal halide perovskites, particularly copper halide variants, have emerged as promising scintillators owing to their large Stokes shift, high photoluminescence quantum yield (PLQY), and efficient radioluminescence. Here, we demonstrate a dioxolane-mediated reaction synthesis as a simple, scalable, and high-yield strategy for producing Cs₃Cu₂Cl₅ powders with near-complete conversion. By employing low-polarity yet water-miscible dioxolane as the reaction medium together with an aqueous H₃PO₂ additive, the solubility imbalance between the precursors inherent to low-polarity solvents is effectively addressed. This approach enables stoichiometry-controlled dissolution–precipitation under a low-polarity reaction environment and the formation of phase-pure Cs₃Cu₂Cl₅ powders with near-unity PLQY. The powders show bright green emission at ~530 nm and retain over 80% of their initial efficiency after 300 days of ambient storage without encapsulation, demonstrating remarkable intrinsic stability. Cs₃Cu₂Cl₅-based composite films deliver an ultrahigh light yield of ~98,000 photons MeV^-1, greatly surpassing that of commercial scintillators such as Gd₂O₂S and CsI:Tl, and exhibiting excellent dose linearity and long-term irradiation stability. When integrated into commercial flat-panel detectors, the Cs₃Cu₂Cl₅-based composite films successfully enabled high-quality 2D radiographic and 3D computed tomography imaging, directly verifying their compatibility with existing X-ray detection systems. This facile and efficient synthesis, together with the demonstrated device-level imaging performance, highlights Cs₃Cu₂Cl₅ as a scalable and practical candidate for next-generation X-ray scintillators.

Abstract

A capillary-type plastic scintillation fiber (PSF) detector combined with a multi-channel analyzer (MCA) was investigated for radionuclide and radiation-type discrimination in in-situ pipe measurements. Plastic scintillators generally exhibit poor energy resolution, resulting in broadened spectra that complicate conventional peak-based identification. In this study, spectral features including the Compton edge (CE), endpoint, and centroid were extracted from β- and γ-ray spectra obtained using Sr-90/Y-90 and Cs-137 sources. The positional dependence of these features along the fiber was evaluated, and significantly reduced variation compared with previous studies was achieved by introducing a reflective structure at the fiber end. In addition, radiation-type discrimination under mixed β/γ conditions was investigated using a stainless-steel (SUS) cover to suppress β-ray components. The results demonstrated that the extracted spectral features remained stable against source position and that condition switching using the SUS cover enabled practical discrimination of β and γ components. These findings indicate that the proposed method is applicable to contamination assessment in confined environments such as piping systems.

Abstract

We are investigating BaF₂:Y scintillators for TOF‑PET applications. BaF₂ exhibits an ultra‑fast scintillation component with a 0.6 ns decay time produced through cross‑luminescence, making it an attractive material for achieving coincidence time resolutions below 100 ps. Recent advances in Vacuum‑Ultraviolet (VUV) SiPMs, which provide high photon‑detection efficiency in the deep‑UV range, have made it increasingly feasible to reliably detect this fast component. In addition to the fast emission, BaF₂ also contains a slow scintillation component with a decay time of approximately 600 ns. This slow component can be substantially suppressed through doping—for example, BaF₂:Y, produced via yttrium doping. Effective suppression of the slow component is essential for applications requiring high count‑rate capability, excellent timing performance, and minimal pulse pile‑up, such as TOF‑PET. In this work, we investigated the coincidence time resolution (CTR) of BaF₂:Y scintillators. The BaF₂ and BaF₂:Y crystals used in this study were grown and polished by the Shanghai Institute of Ceramics (SIC), China. For the BaF₂:Y samples, the yttrium concentration was set to 2 at%, which suppressed the slow component by approximately a factor of four while leaving the fast‑component light yield unaffected. Hamamatsu S13370 VUV‑MPPCs were used to detect the fast scintillation component. Pairs of BaF₂ and BaF₂:Y crystals with dimensions 2 × 2 × 3 mm³ were optically coupled to S13370 VUV‑MPPCs and irradiated with a ²²Na source placed between them in a coincidence setup. We measured CTR values of 86 ps for BaF₂ and 96 ps for BaF₂:Y. These initial results are highly promising, and we anticipate further improvements in CTR with a more optimized experimental setup. Here, we report our preliminary findings.

Abstract

Depth of interaction (DOI) effects can affect the timing response of scintillation detectors due to variations in optical photon transport inside the crystal volume. In this work, DOI-dependent timing measurements were performed using a 1”×2” LaBr₃​ scintillation detector coupled to a photomultiplier tube (PMT) within the framework of the HiSTARs project. A ²²Na source was positioned at different locations along the crystal axis in coincidence with a fast SiPM-based reference detector, and signals were acquired using a DRS digitizer operating at 5 GS/s. The present study focuses on the comparison between interactions occurring close to the PMT side of the crystal and interactions occurring close to the opposite side. Coincidence timing distributions show a systematic shift of approximately 300 ps in the centroid position between both irradiation configurations. Coincidence timing resolutions of 287 ps and 268 ps FWHM were obtained for the two positions, respectively. The observed behavior is consistent with differences in optical photon propagation paths and light collection timing inside the scintillator. Although DOI effects introduce measurable timing shifts, the overall degradation of the timing resolution remains moderate for the studied configurations. Additional irradiation geometries and analysis conditions are currently under study in order to obtain a more complete characterization of DOI effects in long LaBr₃​ scintillation detectors.

Abstract

In order to accomplish rapid and safe decommissioning of Fukushima Daiichi nuclear power plant, the remote and real-time dosimeter was required. To address this issue, detector systems comprising a scintillator coupled to an optical fiber have been proposed.　 In this system, emission wavelength of the scintillator itself should be in the near-infrared (NIR) band, to spectrally discriminate the scintillation signal from the Cherenkov component generated within the optical fiber itself. Therefore, in this study, we focus on Mo doped Cs₂HfCl₆ as the scintillator with NIR emission wavelength and high atmospheric stability. The Mo doped Cs₂HfCl₆ scintillators were grown via melt solidification method.  The result of powder X-ray diffraction pattern suggested that the single phase of Mo: Cs₂HfCl₆ were successfully obtained. Furthermore, the emission wavelength of Mo doped Cs₂HfCl₆ was shifted to 900nm from 420 nm of host emission of C₂HfCl₆. These results indicate Mo: CHC is the promising scintillator for the real time remote monitoring system with optical fiber.

Abstract

Halide scintillators provide a flexible material platform for detectors requiring high gamma-ray energy resolution, fast response, and, in lithium-containing compounds, intrinsic neutron sensitivity. This work presents CrystaLab Co., Ltd.'s compositional-engineering route from LaBr₃:Ce gamma-ray scintillators to Cs₂LiLa(Br,Cl)₆:Ce (CLLBC) mixed-halide Elpasolites for combined gamma-ray and neutron detection. Packaged LaBr₃:Ce detector modules demonstrate high-resolution gamma-ray spectroscopy, with a representative energy resolution of 2.37% at 662 keV and product specifications below 3% at 662 keV. CeBr₃ detectors provide a low-background lanthanide-halide option with energy resolution below 4.5% at 662 keV. By introducing Li and tuning the Br/Cl halide sublattice, Cs₂LiLaBr₆ (CLLB) and CLLBC extend the platform to neutron-sensitive dual-mode detection with neutron-gamma pulse-shape discrimination (PSD). CLLB products provide gamma-ray energy resolution below 4.5% at 662 keV and room-temperature neutron-gamma figure of merit (FOM) values >1.2, while optimized 1-inch CLLBC crystals achieved 3.09% energy resolution at 662 keV and FOM = 2.1. These results show how halide composition, crystal growth, and detector integration can tailor scintillation response from high-precision gamma-ray spectroscopy toward compact mixed-field radiation detectors.

Abstract

Eljen plastic scintillator disks have been coated with layers of neutron sensitiser in a systematic study on the thickness of neutron detection layers. Here we discuss the impact of enhancing standard plastic scintillator (PVT) with coatings containing both lithium and boron isotopes. The application of these layers enhances the material to provide a mechanism for detecting neutron radiation. A systematic study was performed to assess the impact of sensitiser thickness on neutron detection efficiency and neutron-gamma pulse shape discrimination (PSD) performance. Samples were exposed to the mixed radiation fields of a ²⁵²Cf source, and a sample of plutonium. The results demonstrate clear neutron-gamma discrimination in sensitised samples, with measured figures of merit in excess of 3 recorded for optimised lithium-rich coatings. This is comparable performance to established dual-mode detectors. These findings show that thin, externally applied neutron-sensitising layers provide an effective route to enhancing the performance of conventional plastic scintillators. This approach may offer a low-cost means of upgrading existing detector systems in nuclear security applications.

Abstract

We prepared 0.05–1% Cr-doped Al₂O₃ transparent ceramics with a diameter of approximately 10 mm by spark plasma sintering. The samples showed sharp emission peaks at around 700 nm, which matched the sensitivity of typical Si-photodetectors. When the samples were combined with a Si-photodetector and irradiated with X-, α-, and β-rays, currents ranging from picoamperes to nanoamperes were detected. The generated power was maximized when the dopant concentration was around 0.2–0.4 %.

Abstract

An accurate and precise determination of the depth of interaction (DOI) of incident photons in scintillation detectors is paramount to increasing the spatial resolution of positron annihilation tomography (PET) systems. The high number of readout channels is a major source of cost and complexity of such systems, while light sharing multiplexed readouts suffer from edge effects. We have developed a pixelized scintillator detector with DOI capability and 4-to-1 multiplexing utilizing only single-ended readout and light-sharing light guides which can improve signal-to-noise ratio and mitigate edge effects present in other solutions. The scintillator matrix is coupled to a silicon photomultiplier (SiPM) array on one end and light guides on the other. Adjacent crystals are connected in pairs via light guides to enable light sharing. The scintillation light of these crystal pairs is shared between exactly two SiPM pixels. DOI is determined using the ratio of signals on the two SiPM pixels. Pairing distant crystals on matrix edges avoids edge effects. We present the detector architecture and the results of optical simulations evaluating its DOI capability. The interaction position was swept along the central axis of one crystal and the number of incident photons was recorded in the interaction crystal and in the optically connected crystal. The number of rays incident on the sensor coupled to the interaction crystal is always higher than that of the connected crystal. This enables unambiguous determination of the interaction crystal. The ratio of the number of rays incident on the sensors spans three orders of magnitude and approaches unity as the source gets closer to the light guide. These properties demonstrate a favorable DOI capability of the presented detector configuration.

Abstract

The HIE-ISOLDE Timing Array for Reaction Studies (HISTARS) project at CERN is developing a specialized detector system designed to measure the lifetimes of excited nuclear states. For efficient particle detection, thin square GaGG(Ce) are utilized alongside YSO due to their high density and fast scintillation properties. This work evaluates the performance of GAGG(Ce) crystals coupled with two distinct photosensors: a Hamamatsu R9779 photomultiplier tube (PMT) and an array of four silicon photomultipliers (SiPM) using a custom-designed circuit. These configurations were tested in combination with a LaBr₃(Ce) detector to facilitate coincidence measurements. Data acquisition was performed using high-speed digitizer to extract timing and energy information. This article reports on the comparative performance of both readout methods, specifically characterizing pulse shape discrimination capabilities, energy resolution, and the coincidence time resolution for alpha-gamma events. Although the energy resolution is acceptable and it has a good pulse shape discrimination capability, the measured timing limitations show that the current configuration is not ideal for the HISTARS array. These results will guide the selection of better-suited materials for high-precision timing at HIE-ISOLDE.

Abstract

By measuring the time difference and amplitude ratio of the signals at both longitudinal ends of a triangular prism plastic scintillator through dual-end readout, it is theoretically possible to reconstruct the longitudinal position of an incident muon. This paper presents a cosmic-ray muon collimation device that determines the incident region via triple coincidence and features an adjustable incident area. Using natural cosmic-ray muons obtained with this device, dual-end positioning experiments were carried out on a triangular prism plastic scintillator. The threshold-crossing time and amplitude information were extracted from the complete waveforms and fitted, yielding well-fitted relationship functions of the threshold-crossing time difference and the logarithm of the amplitude ratio with respect to the hit position. Experimental results show that the longitudinal position resolution achieved by this method is 3 cm, confirming the effectiveness of the dual-end readout positioning method in triangular prism plastic scintillator array muon detectors and laying a foundation for future research on lightweight and efficient plastic scintillator array detectors.

Abstract

Alkali halide scintillators have emerged as promising low-cost alternatives to conventional thermal neutron detectors such as ³He counters, through the incorporation of neutron sensitive isotopes such as ⁶Li or ¹⁰B, owing to their ease of growth and scalability. In this work, single crystals of CsI:Tl incorporating 2% ⁶Li and varying concentration of Sr as dopants were grown using the Bridgman melt growth method and evaluated for their scintillation and neutron-gamma discrimination characteristics. The radioluminescence spectra are dominated by yellow-green emission from Tl activator sites. Under ¹³⁷Cs excitation, the light output of the crystals was observed to decrease with Sr incorporation compared to the ⁶Li-only doped crystal, while the crystal doped with 2% Sr exhibit the best energy resolution of 6.8% at 662keV. The pulse shape discrimination (PSD) studies of the grown crystals demonstrate effective neutron-gamma discrimination, highlighting the potential of these crystals for use in mixed radiation fields and compact radiation system.

Abstract

A series of polyvinyl toluene (PVT)-based scintillators was prepared via thermal polymerization to investigate the effect of the thermal initiator 2,2-Azobisisobutyronitrile (AIBN) on light output. Samples were synthesized both with varying weight percentages (wt%) of AIBN and without AIBN. All samples contained 2% 2,2-diphenyloxazole (PPO) and 0.05% 1,4-bis(5-phenyl-2-oxazolyl) benzene (POPOP), while the concentrations of vinyl toluene and AIBN were varied. The light output of the developed plastic scintillators (PS) was compared to the commercially available PS EJ-267D using the gamma spectrum from a Cs-137 radiation source. The results indicated that samples containing 0.1% AIBN without argon purging, 1% AIBN without purging, and 0% AIBN with argon purging yielded 24.40%, 28.23%, and 40.88% of the light output of EJ-267D, respectively.

Abstract

TOF-DXA requires simultaneous energy and timing measurement in the low-energy X-ray region (<60 keV).  However, the low X-ray energies result in reduced scintillation light output and small detector signals, making timing measurement challenging. The purpose of this study was to evaluate the performance of different scintillators for energy and timing measurement in low-energy TOF-DXA detectors.

An analog front-end circuit was developed for energy and timing measurement in low-energy TOF-DXA detectors. Detector signals in the energy path were amplified transmitted differentially to reduce common-mode noise and signal degradation. For timing path, a high slew-rate amplifier and comparator were used to generate LVDS preserving the rising-edge integrity of detector pulses. Three 3 × 3 × 3 mm³ scintillators (LYSO, GFAG, and YAP) coupled with a NUV-MT SiPM were evaluated for simultaneous energy and timing measurement. Output pulses were measured using an Am-241 point source to evaluate energy resolution and pulse characteristics. In addition, X-ray spectra were acquired using an X-ray generator operated at 83 kVp and 0.2 mA with a Ce filter to evaluate detector performance in the DXA.

The energy resolutions measured with LYSO, GFAG, and YAP were 25.7%, 24.4%, and 30.9%, respectively. In the timing path, LYSO exhibited the highest slew rate (364.8 MV/s), followed by GFAG (354.9 MV/s) and YAP (220.8 MV/s). LYSO showed the highest peak-to-valley ratios for both the low- and high-energy peaks (3.96 and 1.62, respectively). For the low-energy peak, GFAG and YAP showed PVRs of 3.32 and 2.45, respectively, whereas YAP showed a slightly higher high-energy PVR (1.13) than GFAG (1.02).This overall performance may be attributed to the balanced characteristics of LYSO, which exhibited higher light output than YAP and faster decay time than GFAG.

These results indicated that LYSO was superior to GFAG and YAP for TOF-DXA applications requiring both energy and timing measurement.

Abstract

We report characterization results of a vacuum‑compatible, pre‑production hybrid pixel array X‑ray detector incorporating a silicon sensor and a low‑noise photon‑counting readout ASIC, with emphasis on extending the practical low‑energy discriminator threshold. Calibration pulse measurements combined with refined noise characterization enabled evaluation of discriminator threshold settings down to approximately 450 eV. Considering measured RMS noise levels on the order of 70-80 e^-RMS, such threshold operation implies practical detectability of soft X‑ray photons with energies around 500 eV, although uniform full‑field operation at this threshold is not yet expected.

Low‑threshold operation was found to be temporally stable, showing negligible threshold drift and no observable degradation of pixel operability during continuous operation over extended periods. Soft X‑ray beam validation experiments are planned at the SPring‑8 BL17SU beamline. Additional measurements at higher X‑ray energies will be conducted at BL32‑B2 and BL29XU to support fundamental characterization of detector and ASIC behavior.

The present study focuses on system‑level noise correction and threshold optimization using a conventional silicon sensor. Evaluation with alternative sensor technologies, including iLGAD, is foreseen in future work.

Abstract

We performed a high-energy proton irradiation on a cross-strip germanium detector with the goals of measuring the subsequent loss to detector spectral resolution and characterizing the effectiveness of high-temperature annealing in repairing radiation damage. This work aims to address the expected radiation damage to the Compton Spectrometer and Imager (COSI) during two-year primary mission in low-Earth orbit and evaluate methods for maintenance of spectral resolution. COSI's cross-strip germanium detectors are susceptible to damage induced by the space radiation environment. Negatively charged dislocations induced by high-energy protons trap charge carriers as they drift in the applied bias voltage. The result is a degraded hole-collection signal characterized by widened spectral lines and loss to detector spectral resolution. We irradiated a non-flight detector with a fluence of 1.1 • 10⁷ p⁺/cm² — COSI's expected exposure over prime mission (95% confidence level). After irradiation, we executed a series of high-temperature anneals aimed at physically repairing the damaged germanium. We track spectral resolution by measuring the full width half maximum (FWHM) of the 662 keV spectral line emitted by the isotope Cesium-137. We repaired widened FWHMs through high-temperature annealing to within 12% of its preradiation value, repairing to 2.97 keV compared to the original 2.67 keV. We track spectral resolution across annealing processes and characterize a timescale for repair. We find the anneal time needed to repair damage from COSI's primary mission within 1% of the spectral broadening from residual damage. We project the repair time required at 90°C to be 54 +/- 10 hours.

Abstract

The Compton Spectrometer and Imager (COSI) is a NASA Small Explorer satellite mission scheduled for launch in August 2027. COSI’s imaging capabilities are achieved by reconstructing Compton scattering events within a 2 x 2 x 4 stack of germanium detectors (GeDs). Each GeD contains orthogonal electrode strips on opposite faces of the detector, forming an x-y grid. An incident gamma ray interacts with the GeD by depositing energy in one or more strips on each face that are then read out simultaneously. The “strip pairing” algorithm reconstructs the 2D position of each scatter by comparing the measured energies on pairs of orthogonal strips through a reduced chi square statistic. While simple for a single-site event, the matching becomes more complex when considering multiple scatters within a detector, multiple scatters on a single strip, or charge sharing between neighboring strips. Here, we describe the strip pairing algorithm and how it is implemented into the COSI pipeline. We then test the algorithm's efficacy and use it to demonstrate the imaging and spectral performance of a COSI GeD. Using a ¹³⁷Cs calibration source and a single test detector, we find that the pipeline reconstructs the source's position and yields an energy resolution of less than 1% full width at half maximum at 662 keV.

Abstract

Semiconductor radiation detectors are often limited due to thermal degradation of their electrical properties: as temperature increases, thermally activated carrier generation and other effects lead to elevated leakage currents and reduced signal to noise ratios. For more extreme environments such as advanced reactors and molten salt environments, detectors which operate at high temperatures are needed. Diamond, with its wide 5.47 eV bandgap, high carrier mobility, and radiation hardness is a promising material. Previous work in literature reported charge collection efficiency degradation in diamond at 230 °C, followed by recovery and stability of operation to 450 °C. However, this result has not been widely validated. This study evaluates three diamond samples of varying thickness at elevated temperatures, reevaluating carrier mobility and charge collection efficiency. Alpha traces at each temperature are analyzed to study the current transients, gaining insight into how thermal variations impact detector performance and mechanisms driving charge transport variation.

Abstract

The potential to monitor neutron exposure in hyper-pure germanium detectors via capture-g reactions in the germanium crystal is explored.  The short-lived reactions on ⁷⁰Ge and ⁷⁴Ge resulting in g rays in the 130 to 200 keV energy region have been studied in terms of their response to thermalization of neutrons incident on portable, mechanically cooled, reverse-electrode hyper-pure germanium detectors (REGe^TM). The relative sensitivity in terms of y rays emitted per unit neutron flux is estimated and a comparison is made with the inelastic response observed at higher energies, the latter often considered as cause of damage in such detectors.  We report the potential for these short-lived reactions to serve as a built-in neutron-sensitive diagnostic for germanium detector use in hot cells and realistic, mixed-field environments.

Abstract

This paper quantifies how pixel structural parameters, specifically pitch, photodiode type, and transfer gate design, independently influence neutron-induced white spot generation in CMOS image sensors. Six sensor families with pixel pitches of 0.64, 0.70, 1.00, 1.12, and 1.20 μm, covering single and dual photodiode configurations and single and dual transfer gate designs, were irradiated with neutrons at the TRIUMF facility across seven cumulative fluence steps up to approximately 7×10⁸ n/cm². Ten samples per family were used, with dark-frame measurements performed at each step using eight-frame averaged images. White spot fractions were extracted using a baseline-subtracted delta counting method with Arrhenius-based temperature correction applied to the radiation-induced component. All six families exhibit a clear power-law dependence of white spot fraction on fluence. Spatial analysis reveals a monotonic increase in the 3×3 cluster ratio, defined as the ratio of the summed dark signal increase in neighboring pixels to that of the center hot pixel, from 0.14 at 1.20 μm to 0.24 at 0.64 μm, and this metric remains independent of fluence. A two-stage meta-regression model shows that increasing pitch from 0.64 to 1.20 μm raises the power-law amplitude by 2.12 times, adopting dual photodiode sharing raises it by 2.35 times, and employing a dual transfer gate reduces it by 0.49 times. Furthermore, the model is extended to predict white spot fractions for unmeasured pixel configurations, with potential for radiation-tolerant CIS design.

Abstract

Monolithic silicon CMOS Sensors represent a considerable step forward in charged particle and photon sensing, and emerge as a promising embodiment for the design of novel deep-silicon photon-counting X-ray Computed Tomography scanners . This work presents the first characterization of the ARCADIA MD3 Fully-Depleted CMOS sensors, which technology is currently the baseline for the development of innovative sensors under the framework of the EIC Pathfinder Open 1MICRON project. We report on the characterisation results using an edge-on geometry and a focused synchrotron X-ray beam at the DESY test facility. The x-rays impinge on the edge of the sensor, increasing the active sensor depth seen by the photons, which consequently improves the quantum efficiency with respect to a face-on configuration. A 37 keV focused beam incident parallel to the sensor surface allowed scans along the two directions of the sensor edge, enabling the extraction of the active sensor thickness and mechanical tilt via pixel intensity profiles. Lateral scans across pixel boundaries revealed a periodicity of the spatial modulation of mean cluster size compatible with the pixel pitch of the matrix. Taking the cluster size as the main figure of merit, the performance of the sensor was evaluated at different thresholds and multiple bias voltages in order to assess the variation in the Coulomb cone and collection dynamic. The measurements are supported by Allpix Squared Monte Carlo framework, demonstrating good consistency between the simulated and the real domain.

Abstract

High-energy physics experiments currently use Low Gain Avalanche Diodes (LGADs) as part of their time-of-flight layers to track particles, thanks to their unique temporal resolution and radiation hardness. With increasing deposited charge, the gain was observed to decrease. This gain suppression effect originates from the electrostatic field screening produced by the avalanche multiplication charge. This work investigates the resilience of the LGADs to the gain suppression by characterizing a set of detectors with different gain layer designs and thicknesses, using Transient Current Techniques.

Abstract

Mobility (μ), lifetime (τ), and mobility–lifetime product (μτ) are key metrics for semiconductor radiation detectors. Highly concentrated trapping centers or defects can distort the internal electric-field distribution, producing reduced-amplitude slow-rising charge-sensitive preamplifier pulses that degrade carrier-transit and charge-collection analyses. Although slow-rising-pulse rejection is commonly used to sharpen pulse-amplitude peaks and suppress low-amplitude tails, this work extends its use to charge-carrier transport characterization. A CdZnTeSe detector [1, 2] was characterized at room temperature under electron- and hole-transit-dominant conditions using ²⁴¹Am alpha irradiation. Pulse-shape analysis rejected slow-rising pulses for key metrics extraction. In electron-transit-dominant measurements, the fast-rising pulse events preserved the bias-dependent peak-centroid trend and yielded (μτ)_e = 5.7 × 10⁻⁴ cm²/V, μ_e = 576 cm²/V·s, and τ_e = 1.0 μs, consistent with reported values [3,4]. For hole-transit-dominant measurements, all-pulse events obscured the pulse-amplitude peak and hindered Hecht fitting. After rejection, we obtained (μτ)_h = 2.1 × 10⁻⁴ cm²/V, τ_h = 0.7 μs, comparable to reported values [5], with μ_h = 318 cm²/V·s. Using the rise-time peak as the representative transit time, both all-pulse and fast-rising-pulse events yielded similar μh values. However, when the median rise time of all-pulse events was used instead, drift velocity did not show a clear linear dependence on electric field, whereas fast-rising-pulse events retained linearity. This reflects the need for careful consideration of slow-rising components and the transit-time definition. These results show that slow-rising-pulse rejection enables reliable transport characterization beyond spectral improvement.

Abstract

This contribution presents the capabilities, infrastructure, and technical expertise of the Sensor, Packaging and Integration (SPIN) group at Fondazione Bruno Kessler (FBK). The SPIN unit has established experience in the miniaturisation and integration of electronic systems for radiation detectors, with demonstrated results in space instrumentation (e.g. THESEUS for ESA), high-energy physics (e.g. R&D for ALICE3), and medical applications (e.g. the BOA project within INFN). These domains impose stringent constraints on material budget, geometrical footprint, and signal integrity, motivating the development of dedicated integration strategies. The group core expertise lies in the co-design of sensors and front-end electronics for operation in highly constrained volumes, where standard packaging solutions are not applicable. This includes the optimisation of interconnect technologies (wire bonding, TAB, flip-chip), layout strategies and the implementation of low-noise architectures. Emphasis is placed on preserving signal integrity in high-density environments typical of modern detector systems. SPIN offers end-to-end capabilities covering system design, prototyping, assembly, and characterisation, supported by dedicated infrastructure for precision packaging and electrical testing. A key enabling technology is the use of flexible and hybrid electronics, allowing the development of low-mass, conformable interconnects and non-planar integration schemes. In particular, the group has developed dedicated processes for the fabrication and integration of ultra-thin Kapton-aluminium flexible circuits, optimised for low material budget and high mechanical compliance. These structures enable the realisation of fine-pitch routing on lightweight substrates, with controlled thickness and reduced radiation length, making them especially suitable for tracking detectors.

Abstract

High-purity germanium (HPGe) detectors are widely used in low-background and rare-event measurements because of their excellent energy resolution and relatively high detection efficiency. Compared with a-Ge, sputtered a-Si provides a higher electron-injection barrier, which is attractive for planar detectors under high electric fields. However, the high-voltage (HV) stability of a-Si contacts and the recovery from HV-induced large leakage-current and resolution degradation remain insufficiently understood. The detector was fabricated using a planar p-type HPGe crystal (φ24 mm × 10 mm). A Li-diffused n+ contact was used as the high-voltage electrode, while the opposite face was coated with an a-Si layer deposited by magnetron sputtering. A central Al pad was formed on the a-Si layer by thermal evaporation and used as the signal readout electrode. I–V curves were measured with a picoammeter up to 2500 V, and 57Co spectra were acquired at 500–2000 V using a DSPEC-50 digital spectrometer. The stress test consisted of repeated biasing up to 2000 V, followed by 9 hours of room-temperature storage and 9 hours of thermal treatment with the crystal holder maintained below 70 degrees. After the high-voltage stress test, the leakage current at 1000 V increased to approximately 1.1 nA, more than two orders of magnitude higher than the initial level. After thermal treatment below 70 degrees, the leakage current at 1000 V decreased to less than 2 pA, close to the as-fabricated level. At 1000 V, the resolution broadened to 1.53 keV FWHM, suggesting larger noise and contact-related instability under a higher electric field. After thermal treatment, the resolution at 500 V recovered to 0.67 keV FWHM, comparable to the initial value. In the future, we will repeat stress-recovery cycles, and investigate the dependence on treatment temperature and duration. The possible underlying recovery mechanism related to the a-Si fermi level will also be studied.

Abstract

The COMET Experiment at J-PARC aims to search for the muon-to-electron conversion process with a sensitivity of 10−17. In COMET Phase-I, the muon beam monitor (MBM) will be installed downstream of the primary detector system for background suppression and precise beam control. Since the MBM directly measures the high-intensity muon beam, silicon carbide (SiC), one of the wide-bandgap semiconductors with higher radiation tolerance, is employed as the detector material. The SiC sensor was jointly developed by KEK and AIST, and a beam test performed at J-PARC MLF in 2024 showed successful muon detection and sufficient linearity for operation in the COMET Experiment. In addition, a radiation-tolerant ASIC currently under development at KEK is planned for the SiC detector readout. While the initial prototype MBM board utilised a two-layer polyimide PCB to minimise the material budget, it suffered from external noise and inter-channel crosstalk, obscuring low-amplitude beta-ray signals. To address this and optimise the final MBM board design, we developed a test board equipped with seven SiC detectors. This study showed that noise can be reduced by implementing appropriate noise mitigation techniques, such as a ground plane in a standard PCB structure, enabling stable waveform acquisition of low-amplitude beta-ray signals. In addition, inter-channel crosstalk was effectively suppressed to a negligible level by introducing low-pass filters on the bias supply lines close to each detector. Based on these findings, a prototype MBM board for the final system is currently being developed. This presentation also reports on its design and initial performance evaluation.

Abstract

Lead oxide (PbO) is a high-atomic-number photoconductor for direct-conversion X-ray imaging. This work compares PbO layers produced by three deposition regimes: conventional physical vapor deposition from molten PbO powder, oxygen ion-assisted PVD, and argon ion-assisted PVD. All layers were deposited on indium tin oxide coated glass substrates with a polyimide blocking layer to reduce dark current and were characterized by scanning electron microscopy, Raman spectroscopy, profilometry, density measurements, and X-ray photoconductivity.

Deposition without ion assistance produced low-density polycrystalline PbO with noticeable lag in the X-ray response after irradiation. Oxygen ion-assisted growth yielded dense amorphous PbO with essentially lag-free temporal response. In contrast, argon ion-assisted deposition preserved the polycrystalline structure while producing a more uniform layer with intermediate density and needle-like surface crystallites. X-ray photocurrent measurements showed that the argon-assisted films had an electron–hole pair creation energy, W±, of approximately 6 eV/ehp at 10 V/µm, close to the fundamental limit. Under 30 Hz modulated X-ray irradiation, these films showed stable photocurrent during exposure and rapid signal decay after irradiation stopped, indicating negligible lag.

Abstract

Short-wave infrared photodetectors based on colloidal quantum dots (QDs) have attracted substantial attention owing to their solution processability, tunable bandgaps, and natural compatibility with silicon-based readout integrated circuits. However, the performance of QDs photodetectors is fundamentally limited by the low intrinsic carrier mobility of colloidal QDs, which significantly hinders charge transport. In addition, conventional detector architectures typically provide insufficient internal gain, thereby restricting detection sensitivity, particularly under weak infrared illumination. To address these challenges, we propose a hybrid avalanche photodetector that integrates a halogen-passivated PbS QDs absorption layer with a silicon multiplication region. After absorbing incident infrared photons in the QDs layer, photoexcited electrons are efficiently injected into the silicon multiplication region. The accelerated carriers then induce impact ionization, which enables strong avalanche multiplication and yields high internal gain for efficient carrier collection. Technology Computer-Aided Design (TCAD) simulations indicate that the proposed device delivers a responsivity of 29.5 A/W at a wavelength of 1550 nm, accompanied by fast rise and fall times of 305 ps and 306 ps, respectively. By integrating the spectral tunability of PbS QDs with the robust avalanche gain offered by mature silicon platforms, this architecture provides a highly promising, CMOS-compatible route toward next-generation, high-sensitivity infrared photodetection.

Abstract

Silicon photomultipliers (SiPMs) are attractive photosensors for space applications because of their compact size, low cost, low operating voltage, high photon detection efficiency (PDE), and single-photon sensitivity. Their use in space applications, however, is limited by radiation-induced increases in dark count rate (DCR). This work presents the characterization and radiation tolerance study of FBK NUV-HD-MT SiPMs for the focal plane of the Terzina Cherenkov telescope onboard the NUSES mission, which will operate in a Sun-synchronous low Earth orbit (LEO) at an altitude of approximately 550 km. Devices with micro-cell sizes ranging from 25 to 50 μm were evaluated through a comprehensive characterization campaign comprising static (current-voltage), dynamic (noise and pulse-shape), and optical (photon detection efficiency) measurements to identify the sensor configuration best suited for this application. The 30 μm bare sensors were selected as the optimal solution, achieving a PDE greater than 50% at 400 nm, optical cross-talk below 2%, and a DCR below 60 kHz/mm² at 10 V overvoltage. The radiation environment was modeled using SPENVIS, and dose deposition in the focal plane was evaluated with Geant4 simulations. In parallel, radiation effects were investigated experimentally through two irradiation campaigns: a 50 MeV proton beam test and electron exposure using a Sr-90 source. Proton-induced damage produced a significant increase in leakage current and DCR, while electron-induced degradation was substantially smaller. Temperature-dependent measurements were used to extrapolate in-orbit performance and to evaluate recovery through thermal annealing. The results indicate that performing periodic annealing cycles at temperatures above 50°C can reduce the radiation-induced current by up to 40% after approximately 84 hours, helping preserve detector performance over the three-year mission.

Abstract

To enable the readout of fast signals from SiPM arrays, we have developed a SiPM chip readout based on a compact, low-power radio-frequency preamplifier, thus avoiding the use of ASICs. On a 16 mm² SiPM operating in single photoelectron mode, we measure a rise time of 450 ps and a timing resolution (SPTR) of 84 ps (FWHM), while amplifying the fast component of the SiPM pulses. Furthermore, the active output impedance of the preamplifier is on the order of kΩ, allowing these chips to be mounted along a transmission line without significantly affecting signal propagation. This enables the implementation of dense SiPM lines or arrays, reducing the number of digitization channels required for readout. Finally, an optimized passive pole-zero compensation network, placed at the ends of the line, suppresses the slow component of the SiPM pulses and achieves critical damping of the fast pulse shapes. Thus, pulse pile-up, expected during multiplexing of SiPM photodetectors, can be controlled even at high counting rates. We are currently working on the assembly of such SiPM arrays. We measured a Coincidence Time Resolution of 69 ps (FWHM) using a 9 mm³ LYSO crystal + SiPM spectrometers with our readout electronics.

Abstract

This work presents a real-time FPGA-based Silicon Photomultiplier (SiPM) detector emulator capable of reproducing physically accurate detector waveforms at 16-bit resolution and 2.5 GS/s analog output rate. The proposed architecture combines microscopic event simulation with a temporally quantized event-streaming model, enabling real-time waveform synthesis directly inside FPGA logic while drastically reducing bandwidth requirements compared to traditional waveform streaming approaches. The emulator separates event generation from signal shaping: software layers generate timestamped photon-hit sequences derived either from analytical distributions, Geant4 simulations, or SimSiPM microscopic detector models, while the FPGA reconstructs the analog waveform through parallel Infinite Impulse Response (IIR) filter banks implementing the SiPM transfer function; a look-ahead recursive transformation enables II = 1 operation across all banks at 156.25 MHz, consuming approximately 235 DSP48E2 slices. A multi-phase temporal quantization architecture provides 0.8 ns effective timing precision using a 156.25 MHz fabric clock and eight temporal sub-phases. Experimental validation demonstrates excellent agreement between hardware-generated waveforms and analytical as well as SimSiPM-generated references, including pile-up scenarios with overlapping scintillation events reproduced through linear superposition. The prototype operates with an end-to-end latency of approximately 83 ns.

Abstract

This paper presents the single-cell design and verification of a digital SiPM building block with an integrated temperature-compensation architecture that eliminates external sensors by exploiting SPAD output-amplitude tracking as an intrinsic thermal indicator. A low-power SPAD with a P+/N-well junction and P-well guard ring was fabricated in a 180 nm CMOS process, exhibiting a 10.3 V breakdown voltage, while the digital-cell employs a high-speed comparator to suppress pulse-width variation and premature edge breakdown. The extracted amplitude trend drives an integrated PMIC in a closed loop to dynamically adjust the SPAD bias, maintaining a constant overvoltage and providing instantaneous compensation for thermal drifts.

Abstract

Many experiments rely on measuring vacuum-ultraviolet light, which is emitted in situations as diverse as scintillation in cryogenic liquid noble gases (126.8 nm in argon) and laser spectroscopy of atomic hydrogen isotopes (121.6 nm). Graphene-monolayer photodetectors offer several attractive traits: they are naturally thin, sensitive to a broad range of wavelengths, relatively simple to fabricate (including with integrated readout circuitry with the same process), and can be made on flexible substrates. We present measurements with visible and UV light down to 115 nm, including at low intensity and with fast pulses, of graphene field-effect transistors (GFETs) of various sizes up to 2 mm square fabricated on silicon. These detectors show promise for experiments from neutrinoless double beta decay searches to absolute neutrino mass measurements, as well as dark-matter searches or other VUV-detection applications.

Abstract

Linac Coherent Light Source (LCLS) at SLAC National Laboratory has been going through a major upgrade, called LCLS-II, where the repetition rates are increasing from 120Hz to almost a MegaHertz. Custom ASICs that support x-ray data measurements at these rates are in development for both full frame and sparse data readout. For these ASIC designs we are departing from conventional internal PLL architectures, and instead implementing a 6.25 Gbps serializer with high-speed current-mode logic (CML) receivers (for clock) and transmitters (for data). This approach simplifies the whole chip design, reduces the power consumption and minimizes the noise sensitivity between the high-speed data path and the low-noise front-end blocks. This paper presents detailed performance measurements in a prototype ASIC, followed by the description of their circuit design and implementation.

Abstract

Real time range verification in proton therapy requires analog front end electronics that can process fast, low amplitude signals from scintillating fibers with sub nanosecond precision. This work compares two COTS differential amplifiers (ADA4930 vs. ADA4938) for the NOVO project’s detector. Using actual scintillator derived PWL pulses (0.3 V peak, 18 ns rise time) and a 5th order 250 MHz Butterworth filter, we show that small signal bandwidth governs transient performance. The ADA4930 achieved 660 V/µs pre filter and 350.56 V/µs post filter slew rate, versus 550 and 289.44 V/µs for the ADA4938. Noise analysis reveals the ADA4930 is “resistor noise limited” (61.4% passive resistors), while the ADA4938 is “IC noise limited” (71.6% internal noise). The ADA4930 yields SNR = 65 dB and ENOB = 10.6 bits, outperforming the ADA4938 (62 dB, 10.0 bits). The ADA4930 is therefore the optimal driver for maximizing slope to noise ratio and minimizing timing jitter.

Abstract

The upgrade of the ATLAS Tile Calorimeter readout system for the High Luminosity Large Hadron Collider requires radiation tolerant high throughput electronics capable of reliable long term operation under extreme radiation conditions. This work presents radiation qualification results of the Daughterboard, which interfaces on-detector electronics with off-detector trigger and data acquisition systems. Single point failures are minimized by means of two fully redundant halves, each integrating a CERN radiation tolerant GBTx ASIC for clock and configuration distribution, together with a Kintex Ultrascale FPGA for front-end control and data transmission via SFP+ modules. Radiation qualification targets require tolerance up to 32.71 Gy total ionizing dose, 4.38 × 10E12 n/cm2 non-ionizing energy loss, and 2.37 × 10E11 hadron/cm2 single event effects, including safety factors. Samples of Commercial off the shelf components were tested to evaluate batch variability. Results show all production components meet or exceed requirements. The MOSFETs exhibit degradation via a 15 mV threshold shift after 50 Gy, which does not affect the board functionality. Optical link single event upset error rates to about 9 errors/min over 3584 SFP+ modules, easily manageable by using forward error correction. While 0.65 functional errors/day are recoverable through targeted SFP+ module power cycle. Around 9.37 soft errors/min are expected over 1792 Kintex FPGAs, from which less than 0.20% (1.12 errors/hour) will require targeted FPGA reconfiguration. Radiation effects are mitigated using redundancy strategies and FPGA based soft error mitigation, reducing system impact to negligible levels. The results show that a hybrid architecture combining radiation tolerant ASICs with qualified commercial components enables a robust high bandwidth readout system for sustained operation withing the radiation conditions specified.

Abstract

The new fourth-generation storage rings (diffraction-limited storage rings), such as those at MAX IV Laboratory, ESRF EBS, and Swiss Light Source 2.0, will provide significantly higher brilliance than third-generation sources, resulting in a much larger photon flux at the pixel level. This increased flux will exceed the count-rate capabilities of many single-photon counting detectors currently in use, making them unsuitable for future experiments at these facilities. Matterhorn is a 2-D hybrid pixel detector developed by the PSD detector group at the Paul Scherrer Institute for operation at SLS 2.0. Its main objective is to extend the count rate capability to a target of 20 Mcounts/pixel/s using multi-threshold circuitry implemented in each pixel. The Matterhorn ASIC features a 256 × 256 pixel matrix with 75 μm pitch and is implemented in UMC 110 nm technology. Each pixel includes a charge-sensitive amplifier and a shaper connected to four comparators with independently adjustable thresholds. Setting thresholds above the photon energy enables an extension of the effective count-rate capability. Each comparator is connected to an independently gateable 16-bit counter. The target frame rate for reading out one 16-bit counter per pixel is 10 kHz. To support this rate, the ASIC periphery includes four serial links operating in double data rate mode at up to 3.125 GHz, driven by a 1.5625 GHz clock generated by an on-chip PLL. A fully synthesized digital periphery provides flexible ASIC configuration. Since January 2026, the final ASIC, which is fully functional, has been under test to evaluate its performance. This contribution focuses primarily on the ASIC design and detector architecture, and reports the main performance parameters measured so far, while a detailed discussion of the full characterization results will be presented in a separate contribution.

Abstract

As the scale of the superconducting quantum control system continues to expand, traditional field-programmable gate array architectures face bottlenecks such as high energy consumption, low scalability, and severe signal attenuation. This paper proposes a highly integrated, low-latency digital front-end architecture fabricated in a 28-nm CMOS process to meet dynamic feedback requirements and optimize power, performance,and area. The front-end includes an arbitrary waveform generator and a data acquisition module, independently managed by a microcontroller. At the transmitter, a digital-to-analog converter
is used to output analog signals, with a measured spurious free dynamic range of 52 dB. At the receiver, a low-latency analog-to-digital converter captures analog signals, which the data acquisition module demodulates and ultimately uploads via a low-voltage differential signaling interface. The total hardware latency is restricted to approximately 100 ns. Under full-load testing conditions—including continuous waveform generation,concurrent demodulation of 16 qubits, and state feedback—the total digital power consumption is limited to 539 mW(33.7 mW/qubit). This design achieves extremely low latency and excellent energy efficiency, satisfying the real-time control demands of scalable fault-tolerant quantum computers.

Abstract

Super-low-Earth-orbit (SLEO) optical satellites require compact on-board computing and data-storage electronics, but commercial off-the-shelf (COTS) on-board computers (OBCs) and boot/data-storage media must retain boot, communication, and read/write functions under the mission radiation environment. This paper reports a configuration-level radiation screening of COTS OBC candidates for a 300 km-class, 2-year SLEO optical satellite. A SPENVIS/SHIELDOSE-2 dose assessment gave a total ionizing dose of 14.345 krad(Si) for the design case; therefore, a conservative 30 krad(Si) TID screening target was selected by applying a margin factor of approximately 2. The screening campaign included Arduino and Raspberry Pi family boards, and the detailed design interpretation focused on Raspberry Pi Compute Module 4 (CM4) with input/output board and nonvolatile boot/data-storage media, including Secure Digital (SD) cards, embedded MultiMediaCard (eMMC), solid-state drives (SSDs), and hard disk drives (HDDs). Devices were irradiated with cobalt-60 gamma rays, 10 MeV electrons, and 100 MeV protons, and were evaluated by boot status, communication, supply-current and voltage behavior, read/write operation, and recovery after irradiation. The results indicate that CM4-based configurations were more suitable than the other screened modules for the SLEO OBC candidate. Among storage options, eMMC showed the most favorable radiation and operational suitability, SSDs were identified as a practical secondary storage option after configuration-level verification, and HDDs were not recommended for spacecraft on-board use although some configurations survived to approximately 30 krad(Si), because rotating mechanical parts can affect spacecraft operation. These results support qualifying the exact OBC and boot/data-storage configuration, rather than only the processor board, for COTS-based SLEO on-board computing and storage electronics.

Abstract

The integration of Silicon Photomultipliers (SiPMs) into high-resolution radiation detection systems has become increasingly relevant in the search for novel methodologies in precision energy measurement. This fact implies the need of front-end electronics capable of properly working with these sensors, whose parasitics change among different parts. In the signal processing chain, the initial stage must offer maximal linearity to preserve signal integrity throughout the subsequent architectures. A low-noise goal is also important in order to optimize energy resolution since first stage is critical in noise budget. Consequently, a specialized preamplifier design is presented in this paper in order to address these requirements. The primary design challenge involves achieving a very good linearity through high internal loop gain without compromising system stability. To ensure a controlled frequency response, the circuit employs a Miller compensation scheme whose dominant pole is inside the circuit. Constructing a robust stability topology, the architecture can handle diverse parasitic components in its input due to the use of different SiPM devices. In a first amplification stage, a key innovative structure has been developed as a translinear wideband inner gain stage. Furthermore, to ensure an extended dynamic range which allows working in both photon counting and high amplitude signal conditions without design reconfiguration, a digitally controlled attenuator system is integrated at the output of the first stage. The attenuator uses a R-2R ladder architecture implemented with MOSFET technology, allowing for direct gain adjustments and accurate calibration through external control logic.

Abstract

Hydrothermal volcanic systems are associated with major hazards such as magmatic/phreatic eruptions and toxic gaz emissions.
Monitoring fluid circulation within these systems is a key scientific challenge. 
The SPAGHETTI project (Spectral Panoptic Analyzers of Gamma rays in Hydrothermal Environments for Transport Time Inference) aims to develop
an original instrument dedicated to the continuous and high-frequency monitoring of radon using gamma-ray spectroscopy. 
The instrument relies on NaI scintillation detectors coupled to a specific signal processing electronics. 
The whole system is designed to withstand high temperatures, extreme humidity, and acidic corrosion. 
This contribution focuses on the FPGA-based digital signal processing chain implemented on a low-cost ADC-SoC platform. 
Real-time triggering, digital filtering, and amplitude extraction are performed directly in programmable logic. 
First tests aiming to demonstrate the capability of the system to continuously
acquire and process scintillation signals in conditions representative of volcanic hydrothermal environments will be presented.

Abstract

Performances of tapped delay line (TDL) time-to-digital converters (TDCs) in FPGAs are strongly linked to the FPGA fabrication process, with newer fabrication processes usually yielding smaller propagation delays, thus improving the resolution and precision of the fine measurement. This study evaluates TDLs designed inside the Versal architecture, AMD's newest architecture that uses a 7 nm fabrication process. Since the Versal architecture's new configurable logic blocks (CLB) allow for multiple propagation paths through them, three different implementations of a TDL were evaluated. Code density and single-shot precision measurements were performed on five pairs of channels for each implementation. The lowest quantization error, 4.47 ps, was obtained using a mix of the outputs of the carry propagation circuit and the look-up tables (LUTs) of the CLB, while the best single-shot precision, 7.51 ps, was obtained by using only the output of the LUTs. Because of the change in the CLB design, TDL TDCs perform significantly worse than in AMD's previous architecture.

Abstract

 The Multi-Voltage Threshold (MVT) method is widely employed in particle detection. However, limited by discrete hardware architectures, existing FPGA-based digitization systems struggle to fully exploit the fundamental timing potential of photodetectors. To investigate these performance limits, we systematically evaluate the ultimate timing performance of a detection chain comprising a CMOS SiPM and an in-house-developed MVT ASIC. This ASIC features picosecond time quantization and a wide dynamic range. Experimental results show that, under excitation from a synchronized picosecond laser source, the MVT ASIC achieves an intrinsic electronic time resolution of 16.94 ps (FWHM) and a coincidence time resolution (CTR) of 16.58 ps (FWHM). Benchmarking against a high-performance digital oscilloscope demonstrates that the timing precision of the MVT ASIC closely approaches that of laboratory-grade instrumentation (11.74 ps). This work validates the potential of the MVT ASIC for time-of-flight (TOF) measurement systems and establishes a promising technical foundation for ultra-high-resolution TOF detection.

Abstract

Front-end electronics for nuclear instrumentation require ultra-low noise, high bandwidth, radiation tolerance, and low power consumption to accurately process weak signals generated by radiation detectors. Conventional silicon-based amplifiers face increasing limitations in applications demanding extreme sensitivity, cryogenic operation, and compact integration. This work investigates the use of graphene-based electronic devices as a novel platform for next-generation front-end amplifiers for nuclear science applications. Owing to its exceptional carrier mobility, high thermal conductivity, mechanical flexibility, and potential for low-noise operation, graphene presents a promising alternative material for advanced detector readout electronics.

We present the conceptual design and preliminary evaluation of a graphene-based front-end amplifier architecture intended for integration with nuclear radiation detectors including silicon photomultipliers (SiPMs), cadmium zinc telluride (CZT) detectors, and other low-signal sensing systems. The proposed approach focuses on achieving improved signal amplification while minimizing electronic noise and power dissipation. Device-level considerations including transconductance behavior, impedance matching, frequency response, and low-temperature performance are analyzed to assess suitability for nuclear instrumentation environments.

Theoretical analyses indicate that graphene-enabled amplifier structures can provide enhanced sensitivity and bandwidth compared to conventional semiconductor implementations, particularly in low-temperature and low-signal regimes relevant to modern nuclear physics experiments and radiation detection systems. The results demonstrate the potential of graphene-based front-end electronics as an enabling technology for next-generation nuclear instrumentation, with applications spanning particle physics, medical imaging, homeland security, and advanced scientific sensing platforms.

Abstract

A demonstrator chip integrating a matrix of 32 x 8 readout channels for hybrid pixel detectors has been designed and tested in the framework of the PiHEX project, funded by the Italian Ministry of University and Research. The front-end channels include a low-noise charge sensitive amplifier with detector leakage compensation, an asynchronous, differential comparator, and a 5-bit threshold tuning DAC. A Time-over-Threshold counter is integrated for signal digitization. The front-end channels have been developed in a 28 nm CMOS technology and feature an overall area of 25 um x 100 um with a current consumption close to 4.2 uA. Full details on the design and characterization of the prototype chip will be reported in the conference paper.

Abstract

This paper presents a 5.12-GS/s time-interleaved voltage-controlled-oscillator-based analog-to-digital converter (VCO-based ADC) for high-speed detector waveform digitization in nuclear science and high-energy physics experiments. The proposed ADC employs an 8-channel time-interleaved architecture, where each sub-channel operates at 640 MS/s, resulting in an aggregate sampling rate of 5.12 GS/s. VCO-based quantization is adopted to exploit time-domain signal processing and inherent first-order noise shaping. Each sub-channel integrates a high-linearity bootstrapped sampling switch, a 15-stage ring-VCO quantizer, multi-phase sampling flip-flops, and an asynchronous coarse counter to realize hybrid coarse-fine quantization. To address the major performance limitations of VCO-based and time-interleaved ADCs, a digital calibration scheme is introduced to compensate for the VCO voltage-to-frequency nonlinearity and inter-channel timing skew. The ADC is implemented in a 55-nm CMOS process. Post-layout simulation results show that, at an aggregate sampling rate of 5.12 GS/s with a 2.4-GHz input signal, the ADC achieves an SNDR of 37.5 dB and an effective number of bits of 5.94 bits after calibration. The simulated power consumption is 75 mW from a 1.2-V supply, corresponding to a Walden figure-of-merit of 239 fJ/conversion-step. Full-chip performance measurement is currently underway, and the measured results will be reported at the conference. These results demonstrate the potential of the proposed digitally calibrated TI VCO-based ADC for compact, high-speed detector readout systems.

Abstract

Most high-energy experiments at CERN are upgrad-
ing their trigger systems for the upcoming high-luminosity era of
the LHC. At the same time, development started on the trigger
systems of future collider experiments. To fully validate those
new systems and developments, it is necessary to test both the
algorithms and the underlying interfaces under consideration in a
faithful manner. For the former, simulations are used, and for the
latter small-scale hardware test stands are typically used. This
project proposes the usage of an FPGA-based hardware data
injector, which could replicate at the same time the expected fast
trigger data rate and transmission protocols. By storing Monte-
Carlo simulated data in a DDR4 memory and subsequently
transmitting it continuously to a trigger board under test using
Gigabit Transceivers, the emulator can mimic the expected low
latency data flow.

Abstract

In this article we describe the design and prototyping of the MUlti-channel Streaming Autonomous Readout chip (MU-STAR-mini) to explore the design limits of fast (5 Gsps) sampling array architectures for high voltage resolution, and probe the hypothesis that such architecture can be competitive in resolution and advantageous in terms of power, density and data reduction when compared with traditional ADC architectures for impulsive applications as they are often found in particle physics. We will describe the design and fabrication of the sampling array architecture, the design and fabrication of its test board, and its initial testing.

Abstract

Field programmable gate array (FPGA)-based time-to-digital converter (TDC) has played a very important role in facilities of high-energy physics and scientific instrumentation. Some applications even require TDCs without measurement dead time. In this paper, normal tapped delay line (TDL) based TDC structure is modified towards the realization of zero dead time without losing any TDC original performance. The major challenge in this situation is that there will be more than one signal transition edge in the latched TDL status, therefore, a robust multi-signal transition edge detection circuit module and calibration module are crucial. On a Xilinx UltraScale FPGA, the latched TDL status is first decomposed into 32 divisions, which execute the transition edge picking up parallel before they are combined together to give out all the edges. Since there is only one calibration table in the system, all the edges have to be calibrated sequentially through a specially designed data queue. For performance evaluation, the TDC prototype is implemented on the FPGA with a system clock frequency of 700 MHz. The TDC resolutions for the rising and falling edges are measured as 2.44 ps and 2.31 ps respectively, meanwhile the root-mean-square (RMS) time-interval measurement precisions are measured in the range of 3.5 to 6.5 ps. Applying two trigger signals with 500 ps separation into the TDC repeatedly, all the edges are captured and correctly measured, which demonstrates the capability of zero measurement dead time of the design.

Abstract

The Circular Electron Positron Collider (CEPC) is a large-scale international high-energy physics facility with the Higgs boson as its primary scientific goal. Three operation modes are foreseen: Higgs, Z, and W+W-. The CEPC Trigger and Data Acquisition (TDAQ) system faces a fundamental challenge: extracting rare physics events from an extremely high background rate. For example, in the Higgs mode, simulation results indicate that the target physics event rate is approximately 0.5 kHz, while the bunch crossing rate reaches 1.34 MHz. To cope with this challenge, the required Level-1 (L1) trigger rate for the Higgs mode is about 20 kHz.This extremely low signal-to-background ratio demands a trigger system capable of making efficient and real-time trigger decisions to achieve high-purity event selection while maintaining high efficiency for rare physics signals.To facilitate research and verification of the trigger system design, this talk presents a general-purpose Advanced Telecommunications Computing Architecture (ATCA) processing unit for the CEPC TDAQ system. The unit features an AMD Virtex UltraScale+ field-programmable gate array (FPGA) as its computing core, providing high-density parallel processing. High-speed data transmission is achieved via the FPGA's Gigabit Transceiver (GT) and optical transceivers, offering standardized interfaces for electronic subsystems. The design incorporates 11 Quad Small Form-factor Pluggable (QSFP) modules providing 44 high-speed lanes up to 25 Gbps each, 16 GB of Double Data Rate (DDR) memory for hundreds of milliseconds of buffering, and an Intelligent Platform Management Controller (IPMC) + K26 system-on-module (SOM) for intelligent management. Developed as a flexible testbed, this unit enables the implementation and evaluation of various trigger algorithms and architectures. The test results of related trigger functionalities and performance will be presented in the talk at the conference.

Abstract

The China Spallation Neutron Source (CSNS) is currently upgrading to CSNS-II, increasing the proton beam power to 500 kW and constructing several new neutron spectrometers. The CSNS-II upgrade poses significant challenges for detector and electronics systems, including higher counting rate, improved position resolution requirement, and vacuum operating environment. The large-area ³He tube array detector has been adopted as the major neutron detector owing to the high detection efficiency, the excellent neutron/gamma discrimination, and the ability to construct the detector with large area coverage. In this paper, we present an ultra-low-power front-end readout electronics designed for ³He tube detectors in CSNS-II spectrometers. The circuit design, FPGA algorithms, and laboratory test results of the electronics will be discussed. Experimental results demonstrate that the new electronics system can achieve a single-tube counting rate of 100K/s while maintaining a position resolution about 6 mm. Test also confirmed the system's capability to operate under vacuum conditions. These results meet the requirements for neutron signal readout in CSNS-II spectrometers.

Abstract

This paper presents a modular, per-channel self-triggered firmware architecture for the FPGA-based readout of electromagnetic calorimeters at the INSIGHT experiment. The firmware integrates a compact deconvolution stage to compensate preamplifier trailing-edge ringing, a configurable, pipelined feature extraction chain (baseline estimation, peak-based trigger with interpolation, pileup detection), and a lightweight network module. The implementation uses a decomposition of the deconvolution kernel to reduce resource usage. Despite the computational demands, the implementation of the deconvolution is resource-efficient and leaves headroom for further extensions. A Python-based simulation framework guided the selection and tuning of algorithms and numerical parameters. Test-beam measurements with a detector module showed an effective ringing suppression without affecting the pulse amplitude and timing; cooling the detector to the nominal operating temperature in upcoming runs is expected to further reduce residual artifacts. The resulting firmware framework provides a base for both electromagnetic calorimeters, the Crystal Barrel and the PANDA forward endcap calorimeter, in the INSIGHT experiment and will reduce maintenance effort.

Abstract

The Polarimetry Focusing Array (PFA) onboard the planned enhanced X-ray Timing and Polarimetry (eXTP) observatory comprises three identical X-ray imaging telescopes, providing X-ray polarimetry in the 2–8 keV energy range with high spatial, spectral, and temporal resolution. To meet these requirements, each detector unit integrates a radiation sensor with a dedicated Application-Specific Integrated Circuit (ASIC) for precise charge extraction and digitization of individual X-ray events.

This paper presents a readout ASIC designed for this application. The proposed ASIC features a self-triggering mechanism that enables event-driven readout with dynamically configurable windows. To satisfy stringent performance requirements, a two-stage analog readout architecture is adopted. Although the target system employs a 352 × 304 pixel array, a prototype ASIC with a 70 × 60 pixel matrix and a pitch of 50 μm was fabricated in a 180-nm CMOS process to validate the design.

Measurement results demonstrate an Equivalent Noise Charge (ENC) below 40 e⁻ rms. With a 12.5 MHz readout clock, the dead time for a typical event involving approximately 1000 pixels is maintained below 100 μs. These results verify the feasibility of the proposed ASIC architecture for the eXTP PFA instrument.

Abstract

Cerium doped gadolinium aluminium gallium garnet (GAGG:Ce) is capable of detecting both photons and charged particles. Particle identification capabilities present GAGG:Ce as a useful detector in nuclear physics research, and in medical, industry, security and space applications. This work explores the feasibility of using trapezoidal filtering for pulse shape discrimination (PSD) to perform particle identification so that the pulse height analysis (PHA) mode in commercially available waveform digitizers can be used without developing custom PSD firmware. Waveforms recorded from proton and alpha scattering measurements were analyzed offline to optimize the trapezoidal filter parameters. Satisfactory pulse shape discrimination was achieved with a figure of merit of 0.68. Future work will incorporate the optimized parameters in the waveform digitizers for heavy-ion reaction experiments to improve particle identification in real time.

Abstract

The HIE-ISOLDE Timing Array for Reaction Studies (HISTARS) is being developed for the measurement of lifetimes of excited nuclear states produced in reactions at HIE-ISOLDE (CERN). It is based on fast particle and gamma scintillator  detectors that will require a fast data acquisition system capable of handling more than 60 channels with good time and energy resolution. Digitizing the fast scintillator signals will allow for a flexible signal processing, while preserving the pulse time structure, and potentially enabling high-precision timing.
In this work we measure the intrinsic time resolution of several digitizer systems for data acquisition and processing of data stemming from fast scintillators, such as LaBr$_3$(Ce). A very promising high-performance candidate is the CAEN 2751 digitizer, providing 14-bit, 1 GS/s, 16 channels, and embedded FPGA processing for advanced waveform acquisition and digital pulse processing. 
The DRS4 Evaluation Board uses the DRS4 switched-capacitor array chip developed at Paul Scherrer Institute with 4 input channels, sampling rates of 0.7 to 5 GS/s with 1024 sampling cells at 14-bit resolution. The CAEN DT5742B is a 16-channel switched-capacitor digitizer based on the same DRS4 chip, providing 14-bit resolution and up to 5 GS/s sampling rate.
The CAEN DT5751 is a desktop digitizer 4 channels, 10-bit resolution, and sampling rate of 1 GS/s per channel.
An in-house custom microcontroller-based data acquisition system, capable of analog pulse integration and timestamp assignment based on a bespoke time to amplitude converter (TAC), is also included in the DAQ systems comparison. These results will guide the choice of the data acquisition system for the HISTARS project and determine its baseline contribution to the timing capabilities of the array.

Abstract

Silicon Drift Detectors (SDDs) are the most widely used sensors for energy dispersive X-ray detection in the energy range from a few hundred eV to some tens of keV. When connected to a state-of-the-art CMOS preamplifier, they provide high energy resolution at short peaking times, allowing for high count rates. In high-count-rate X-ray fluorescence (XRF) measurements, SDD arrays are also considered; however, using conventional Digital Pulse Processors (DPPs) requires a large number of ADCs and FPGAs, leading to an increase in overall system size. To keep the total system compact, this study proposes an approach using the analog pulse processing ASIC "SCARLET" and investigates its feasibility. Using an actual XRF setup, a single SDD signal was input into the ASIC as an initial evaluation of its fundamental throughput performance and energy spectrum at an input count rate of up to 1.8 Mcps. The results show that with a 400 ns shaping pulse width, the ASIC could potentially achieve faster throughput compared to a conventional DPP with an equivalent filter length. However, with a 200 ns pulse width, the throughput decreased, and incorrect energy measurements were observed, such as the appearance of an unexpected peak on the lower energy side. This indicates that there are remaining issues in the peak detection logic within the ASIC. This study provides fundamental insights into the effectiveness of the analog ASIC for high-count-rate XRF measurements and highlights areas for future improvement.

Abstract

Timepix4 represents a significant advancement in the Timepix family of hybrid active pixel detectors, particularly attractive due to its fine time resolution (195 ps) and high supported hit rates (up to 5 GHz/mm2). To address performance challenges presented by these new features, an upgrade has been implemented for the Track Lab software package, which has become a standard tool for data collection and analysis of many
Timepix-based systems. Above all, the presented upgrade adds support for a 40 Gb/s PCI-express readout interface, capable of reaching hit rates of 300–400 MHz. To facilitate efficient analysis of such data volumes, the presented work implements a number of optimizations, including memory-level bit packing, deduplication and GPU offload for hit clustering, a required step for most event-based analysis algorithms. This work will
present details of these improvements, characterize Track Lab’s response under heavy load and present results of a performance benchmark. The upgraded software is expected to find plentiful applications in high-intensity X-ray imaging, γ-imaging and the forthcoming high-luminosity campaign at the Large Hadron Collider.

Abstract

This paper presents a scalable and high-throughput data acquisition (DAQ) architecture developed for real-time proton beam monitoring applications in advanced proton therapy systems. The proposed system is designed to support detector-based tracking of proton energy deposition, enabling continuous monitoring of beam delivery and allowing treatment parameters to be adjusted dynamically during operation. By providing precise localization of proton energy release within the target region, the system contributes to improved dose accuracy while minimizing radiation exposure to surrounding healthy tissue. The DAQ architecture integrates high-speed digitization, synchronized timing distribution, and efficient event aggregation within a modular rack-based platform optimized for detector readout applications. The system consists of eight digitizer boards supporting 64 analog input channels corresponding to 32 detector elements. Differential signal readout is employed to improve signal integrity and reduce noise susceptibility during acquisition. A dedicated Data Collection Board equipped with 40 Gbps QSFP+ communication links aggregates and transfers acquired event data to a host DAQ server with high bandwidth and low latency. System-wide synchronization is maintained through a centralized Clock Distribution Board, while an OpenVPX-compliant Power Supply Board ensures reliable operation of all modules. The proposed architecture provides a compact, scalable, and synchronized solution for real-time proton therapy monitoring and high-performance medical imaging applications.

Abstract

The Medipix4 ASIC is a hybrid pixel detector developed for high-rate spectroscopic X-ray imaging and photon-counting applications. Its synchronous frame-based architecture supports multiple acquisition modes and up to sixteen DDR output links, requiring scalable off-detector electronics. This work presents a modular FPGA-based Data Acquisition (DAQ) system developed for Medipix4 characterization and detector integration.
The system is composed of a Xilinx Kintex UltraScale FPGA DAQ board and a dedicated CERN-developed Medipix4 chipboard directly interfaced through FMC connectivity. The firmware reuses the IPbus-based control path developed for the Timepix4 ASIC while implementing a dedicated Medipix4 synchronous acquisition pipeline. Sixteen independent DDR interfaces operate at selectable clock frequencies of 20, 40, 80, 160 and 320 MHz. Each detector link is connected to a deserializer implementing frame detection, header decoding and 64-bit payload reconstruction. The decoded headers allow the firmware to determine the active acquisition configuration, including counter depth, active links and payload size.
The sixteen parallel data streams are packetized and transmitted through two independent 10~Gigabit Ethernet UDP links. The aggregate detector bandwidth of 10.24 Gb/s is compatible with the measured sustained transmission throughput of 2 X 8.85 Gb/s. The firmware uses reusable open-source HDL components where possible and is distributed within the Medipix collaboration as Open Hardware under the CERN Open Hardware Licence. Ongoing developments include migration toward lower-cost FPGA platforms and support for simultaneous acquisition from multiple Medipix4 ASICs.

Abstract

We measured energy-dependent single-event-upset (SEU) rates for three field-programmable gate array (FPGA) types under neutron irradiation at the BL10 beamline of the Materials and Life Science Experimental Facility (MLF) at the Japan Proton Accelerator Research Complex (J-PARC). Two 28-nm complementary metal oxide semiconductor (CMOS) devices and one 65-nm CMOS device was irradiated with neutrons over a wide energy range from meV to MeV. Dedicated firmware detected SEUs with high timing precision, and the time-of-flight (TOF) method identified the neutron energies that caused the SEUs. We estimated the neutron fluence from a PHITS simulation of the beam spectrum and normalized the SEU counts to the fluence for each FPGA type.

The SEU rates clearly depended on neutron energy. 
One 28-nm device had an SEU rate in the 10^-8-10^-7 MeV region about one order of magnitude higher than that at 1 MeV, whereas the other devices showed no significant low-energy enhancement. Below 0.1 MeV, the SEU rate increased as neutron energy decreased, approximately following the 1/v dependence, where v is the neutron velocity.

The two 28-nm devices responded differently despite using the same process technology. In contrast, the 65-nm device had a low-to-high-energy change similar to that of one 28-nm device.
These results suggest that device structure or neutron-sensitive materials, rather than process technology alone, contribute to neutron-induced SEUs.

Abstract

This work presents the design and characterisation of the front-end eletronics of a wireless active personal dosimeter for interventional radiology. The device is a low-cost, single-channel sensing unit that can be replicated across the thorax, neck, hands, and eyes, providing operators with continuous dose-rate feedback at these four anatomical regions. Each dosimeter consists of a tissue-equivalent plastic scintillator coupled to a silicon photomultiplier, read out by a charge-sensitive amplifier, a semi-Gaussian shaper with pole-zero cancellation, and a peak-and-hold circuit triggered by a comparator. Power management and wireless transmission are handled by a battery-operated microcontroller, keeping the assembly compact and unobtrusive to the operator. Dose-rate data are transmitted to a remote display with visual and sound alarms. The front-end has been characterised under dark-count conditions, with a dead time of 11.6 microseconds expected to extend to approximately 20 microseconds once analog-to-digital conversion is included in the readout loop. Energy and dose calibration of the coupled detector, together with the four-channel graphical interface, are the next development steps.

Abstract

This work presents the development of a versatile FPGA firmware for the readout of ³He resistive wire tubes widely used in neutron scattering instruments. These detectors are deployed in large numbers across major neutron facilities, including ISIS, ILL, SNS, FRM2, ESS, MLF, and CNCS.

In typical configurations, detector signals are first conditioned by front-end preamplifiers located near the tubes, then digitized by ADC systems positioned up to 40 m away. The FPGA-based processing extracts pulse amplitudes to determine the neutron interaction position along the detector. The implemented system is based on the R5560 ADC, operating at a sampling rate of 125 MHz.

The firmware, developed using SCI-Compiler, is organized into two main blocks: signal processing and data transmission. The signal processing stage includes baseline correction and gain scaling to maximize ADC dynamic range, followed by configurable filtering. The latter consists of cascaded moving-average integrators and a differentiator, with user-defined parameters (number and length of filters) adjustable via control registers to accommodate different instrument requirements.

To ensure compatibility with both legacy and next-generation data acquisition systems, the firmware supports multiple communication interfaces. At ISIS, the ADC serial interface is currently used for integration with the existing Data Acquisition Electronics (DAE), as demonstrated on the OSIRIS instrument. In parallel, a high-throughput UDP-based optical fibre link has been developed for future deployment, including WISH and μRWELL-based detector systems. Additionally, a TCP/IP Ethernet interface provides a flexible solution for diagnostics and lower-rate data acquisition.

This architecture enables a scalable and adaptable readout solution for current and future neutron detector systems.

Abstract

The Light Dark Matter eXperiment (LDMX) is a proposed fixed-target experiment that will search for sub-GeV dark matter using a missing-momentum signature in high-rate electron interactions. Its Silicon Tagging Tracker (STT) and Silicon Recoil Tracker (SRT) measure incoming and outgoing charged particles around the target, providing trajectory and momentum information needed to identify events with invisible final-state particles and reject backgrounds. The Tracker data acquisition (DAQ) system is built around APV25-based silicon strip modules, front-end boards (FEBs), and a Bittware PCI Express back end integrated with the common LDMX timing, fast-control, and event-building infrastructure. Each FEB distributes APV25 clock, trigger, reset, I2C control, low-voltage power, and sensor bias. It also digitizes APV25 analog streams, formats three-sample channel records, and transmits Tracker event records to the back end. The architecture is constrained by the APV25 pipeline latency and must preserve bunch timing while scaling to the final multi-FEB Tracker. During the 2025 LDMX test at SLAC End Station A, a Tracker DAQ vertical slice operated one FEB with four silicon Hybrid modules and the Bittware back end. The system acquired coherent data with other LDMX subsystems, demonstrated stable noise performance, was timed in to beam, and observed beam-induced hits with the expected timing behavior.

Abstract

The Neutron Veto (NV) of the XENONnT experiment is a gadolinium-doped water Cherenkov
detector surrounding the central xenon time projection chamber (TPC). It is instrumented with 120
photomultiplier tubes (PMTs) to reduce the nuclear recoil background induced by radiogenic
neutrons originating from detector materials, which can mimic WIMP-like signatures.
During the third science run (SR2), conducted between October 2023 and March 2025, XENONnT
achieved a significantly reduced background configuration. In particular, the neutron background
was suppressed by loading the water tank with 500 ppm of gadolinium sulfate, resulting in an
approximately twofold reduction compared to earlier runs performed with pure water. Following
maintenance activities initiated in summer 2025, the detector is currently being recommissioned in
preparation for the next science run.
This work presents the latest upgrades and the operational performance of the NV data acquisition
(DAQ) and online processing systems. The NV DAQ is based on a fully triggerless architecture in
which PMT waveforms are continuously digitized, enabling software-based coincidence searches
within time windows of several hundred nanoseconds.
To enhance long-term reliability and stability, the DAQ system has been upgraded with new-
generation CAEN PCIe readout cards. Concurrently, the front-end server has been upgraded to
ensure compatibility with latest operating systems, while improving fault tolerance, system
monitoring, and integration with the experiment-wide online processing system.
A further improvement is the deployment of a dedicated FPGA-based monitoring board for the NV
data readout. This board processes digitizer monitoring signals such as buffer occupancy, busy and
flow-control flags, timing and synchronization signals, and error conditions during both science
data taking and calibration runs, enabling continuous checks of DAQ performance and stability.

Abstract

Lithium Indium Phosphorus Selenide (LiInP₂Se₆) has emerged as a promising semiconductor material for neutron detection due to its high neutron capture cross-section, suitable bandgap, and favorable charge transport properties. This work reports on the development of a modular LiInP₂Se₆ detector prototype for single-crystal diffractometry (SCD) applications. The prototype consists of an 8 × 2 array of LiInP₂Se₆ strip sensors with a 300 μm pitch configuration integrated onto a custom sensor baseboard. Initial sensor characterization includes electrical IV measurements and preliminary radiation-response evaluation. The sensors are coupled to a dual-ASICs daughterboard based on VATAGP7 ASICs for signal readout and processing, while detector control and data acquisition are managed through the LISeDAQ acquisition system. This work demonstrates the integration of LiInP₂Se₆ sensors, front-end electronics, and acquisition hardware into a scalable detector module architecture toward future high-resolution SCD detector systems.

Abstract

In airborne radiation monitoring using unmanned aerial vehicles conducted in Japan following the Fukushima Daiichi nuclear power plant accident, a flat-source model is used to estimate the ambient dose equivalent rate at ground level from count rates. However, its assumptions are not suitable for mountainous and forested areas. To improve reliability in these areas, an artificial neural network was constructed that incorporates topographic and vegetation information as input variables in addition to count rate information, and its impact on estimation performance was evaluated.

The ambient dose equivalent rate map generated by the artificial neural network model reproduced spatial variation in more detail compared to the map created by the flat-source model. Vegetation information helped distinguish between forests and open land, contributing to improved estimation accuracy and precision. Topographic information reduced errors in mountainous areas where the relative position of the detector and the ground changes drastically. An artificial neural network incorporating this information can contribute to identifying hotspots and determining whether evacuation orders can be lifted for small areas.

Abstract

Machine learning has shown strong potential for radionuclide identification using gamma-ray spectroscopy. However, its black-box nature limits the physical interpretation of model predictions. In this study, a spectroscopy-tailored explainable artificial intelligence (XAI) framework for a 1D-CNN-based radionuclide identification model was investigated using gamma-ray spectra measured with a NaI(Tl) detector. The model was trained to classify four radionuclides: ²²Na, ⁶⁰Co, ¹³⁷Cs, and ⁵⁴Mn. To identify the spectral regions contributing to the CNN predictions, Grad-CAM, LIME, and Kernel SHAP were applied. In particular, because perturbation-based XAI methods are strongly affected by the definition of input features, two spectral banding strategies were compared: conventional uniform banding and FWHM-aware adaptive banding. The results showed that compared with uniform banding, the proposed adaptive banding provided more localized and physically interpretable attribution results in several cases. These findings suggest that detector-resolution-aware banding can improve the reliability and interpretability of XAI for gamma-ray spectroscopy.

Abstract

This work demonstrates source-transferable neutron-gamma pulse-shape discrimination in 3D-printed plastic scintillators using compact deep neural networks trained on experimental waveforms. Additively manufactured plastic scintillators are attractive for application-specific neutron detectors because printing can produce nonplanar and architected geometries. However, their pulse shapes can depend on resin chemistry, dopant loading, and curing conditions. Therefore, discrimination algorithms validated on conventional liquid or cast plastic scintillators cannot be assumed to transfer directly to printed scintillators. Pulses were collected from ²⁵²Cf and AmBe sources using CAEN DT5751 and DT5730 digitizers, respectively, and analyzed using charge-comparison pulse-shape discrimination, unsupervised clustering, tree-based classifiers, one-dimensional convolutional neural networks, and fully connected deep neural networks. Charge-comparison methods were limited to 92-93% accuracy for the more overlapping ²⁵²Cf dataset, whereas the optimized network used smoothed, peak-aligned, amplitude-normalized waveforms and achieved 99.55% test accuracy on ²⁵²Cf. Without retraining, the same model achieved 99.91% inference accuracy on the independent AmBe dataset. Peak alignment was found to be essential as AmBe inference accuracy increased from 88.11% for unaligned raw waveforms to 99.91% for aligned waveforms. These results establish peak-aligned waveform learning as an effective route for robust neutron-gamma discrimination in 3D-printed plastic scintillators.

Abstract

Gamma-ray anomaly detection and identification in complex environments requires background models that are accurate, robust to environmental variation, and interpretable enough to support adaptation given contextual information. We present Gaussian Process Prior NMF (GPP-NMF) spectro-temporal background modeling, combining the physically meaningful structure of Non-negative Matrix Factorization (NMF) in the spectral dimension with a Gaussian process prior (GPP) in the temporal dimension, parameterized by a smoothing hyperparameter λ_T. Evaluated on RADAI synthetic mobile search data, GPP-NMF outperforms standard spectral NMF in the low false alarm rate regime of operational interest, with a single fixed calibrated λ_T. We validate a principled approach for λ_T adaptation to contextual information by showing that the power spectrum of the logarithm of the total building solid angle seen by the detector — extractable from camera data in real time — closely matches the GPP prior power spectrum. Full contextual integration results, including on real-world data, will be presented.

Abstract

The primary objective is to describe and evaluate a multisensor architecture in which several small unmanned aircraft jointly support geographically registered ambient gamma survey without carrying a separate heavy scintillator and high-resolution pixel imager on every vehicle. The working hypothesis is that conventional handheld Geiger–Müller survey instruments, rigidly mounted with the numeric display visible to an onboard camera, can supply useful count-rate estimates when digits are recovered by image-space region selection, classification or template matching, and short-window temporal filtering that suppresses vibration-induced blur. The research design couples this optical readout path to a single mission manager that maintains a unified operator display, explicit multicopter formations to spread measurement locations along transects or grids, leader–follower roles with reassignment when radio links degrade, and a fixed policy in which facial-identity checks gate escalation to autonomous rerouting or alarm acknowledgement. Methods include hardware-in-the-loop bench comparison against an independent counting reference, latency measurement on embedded accelerators, and qualitative stress tests for glare, backlight change, and partial occlusion from neighboring rotors. Preliminary outcomes indicate stable decode at representative dose rates when temporal aggregation spans one to two seconds, inference on the order of tens of milliseconds per frame region, and graceful degradation when the display is shadowed. The conclusions are that display-mediated readout is a viable interface when spectroscopic identification is not required, and that formation metadata should be fused with decoded rates before mapping. Implications are outdoor line and grid flights with uncertainty accounting, handheld references, and optional spectrometer comparison before field deployment claims.

Abstract

Non-intrusive inspection of cargo containers via dual-energy X-ray radiography increasingly faces threats concealed within the container’s own structural void spaces, making segmentation of the structural region a necessary prerequisite for downstream anomaly detection. Two compounding factors hinder this task: inconsistent geometric distortion across operational scans, and path-integral structure–cargo overlay that prevents pixel-level human annotation in densely loaded regions. We formulate the task as binary segmentation and propose a regionwise generative overlay augmentation pipeline, in which Stable Diffusion is adapted via a three-stage scheme—LoRA, ControlNet, and inpainting—to synthesise structure–cargo overlay imagery. Polygon annotation is partitioned by visual confidence: structural polygons supply segmentation labels and ControlNet conditions, while cargo polygons serve as inpainting masks. The central hypothesis is that a model trained with the proposed structure–cargo overlay augmentation outperforms a baseline trained on real annotations alone, evaluated on two test partitions defined by the visual clarity of structure–cargo separation under overlay: Easy (clearly distinguishable) and Hard (separation challenging but feasible).

Abstract

Reliable gamma-ray spectrum databases are needed for data-driven radioisotope identification, but direct measurements of accident-related or short-lived radionuclides are often limited by safety regulations and isotope availability. Geant4 can generate synthetic spectra, but simulated spectra may differ from measured detector-domain spectra because of detector response, energy resolution, gain variation, and channel-domain distortion. In this study, we evaluated a U-Net-based sim-to-real translation framework to transform Geant4-simulated spectra toward measured detector spectra. Measured detector energy-resolution broadening and channel-domain augmentation were applied before translation, and a spatial-transformer-integrated U-Net was used to compensate for gain- and offset-related channel shifts. In validation, the model achieved a mixed-isotope training mean absolute error of 0.045214 and leave-one-out mean absolute error values of 0.068125 for unseen ²²Na and 0.047816 for ¹³⁷Cs. A photopeak-based fidelity analysis showed that 44% of ²²Na samples and 24% of ¹³⁷Cs samples satisfied the defined peak-position and peak-height criterion and a peak-height difference threshold of 0.05. These results indicate that U-Net-based translation can partially reduce the sim-to-real discrepancy in gamma-ray spectroscopy, while consistent reconstruction of detailed photopeak structures remains limited.

Abstract

This work addresses real-time identification of gamma-emitting nuclides released outside a damaged reactor using a network of low-cost Silicon photomultiplier (SiPM)–scintillator gamma detectors. Of all fission products, only a few reach the plant exterior in amounts sufficient to produce measurable photopeaks; using the NUREG-1465 source term for Pressurized Water Reactor (PWR) severe accidents, we reduce the monitored set to the dominant nuclides ⁸⁵Kr, ¹³¹I, ¹³²Te, ¹³⁷Cs, ⁶⁰Co, and ¹⁴⁰Ba. For this set, we design a 2K+1 physics-region preprocessing front-end that places one region of interest on each target photopeak and Compton edge, producing compact features for small downstream classifiers, including the target on-chip deep-learning core. Because direct validation on a real off-site release is impractical, we evaluate the pipeline on a K=5 surrogate set of ¹³³Ba, ²²Na, ¹³⁷Cs, ⁵⁴Mn, and ⁶⁰Co spanning the same 80 keV-to-1.3 MeV band on a NaI(Tl)-SiPM detector. On this surrogate, the 2K+1 preprocessing achieves 83.9% accuracy at 25 photopeak counts per window and 97.8% at 100, compared with 75.4% and 94.3% for a raw pulse-vector baseline.

Abstract

Vehicle-based radiation monitoring systems are widely used for environmental radiation surveillance, nuclear emergency response, and radioactive contamination mapping. Conventional systems use large-area NaI(Tl) scintillation detectors with high sensitivity for low-count-rate radiation measurements. However, they have limitations such as poor energy resolution and the inability to provide directional information of radiation sources. To overcome these limitations, this study developed a vehicle-based radiation source direction detection system using nine NaI(Tl) detectors while maintaining high sensitivity and energy resolution. The system consists of nine Φ3” × 3” NaI(Tl) detectors. Four upper-layer detectors were arranged for rapid direction identification, one center detector was used for dose assessment, and four lower-layer detectors with lead collimators improved directional accuracy. Radiation source direction was estimated using detector count-rate ratios according to incident direction and distance. An ANN-based algorithm was applied for direction estimation. System performance was evaluated using a 137Cs (100 μCi). Energy resolution and sensitivity were evaluated at 23 cm, while angular resolution tests were conducted at 6 m with source angles varied from 0° to 90° at 10° intervals. The developed system showed approximately 6.8 times higher detection sensitivity than a conventional large-area NaI(Tl) detector system. The angular resolution was evaluated as 45° and 10° for the upper and lower detector layers, respectively, confirming the feasibility of radiation source direction detection in a vehicle-based system. Future work will focus on miniaturization through compact optical sensors, detector rearrangement, and signal-processing optimization.

Abstract

Detector networks are a potential method of surveillance of radiological material moving through urban environments. Such a network would require ways of automatically assessing collected spectra for non-background isotopes, particularly in count-starved spectra collected in environments with complex backgrounds. In 2018, the London-based SIGMA trial gathered 1.5 billion one-second gamma spectra using both stationary and mobile scintillator detectors. This work presents a wavelet analysis method optimised for use with count-starved gamma spectra. A range of ¹³⁷Cs spectra were collected, both experimentally and simulated in Geant4, and combined with real-world SIGMA background data. A continuous wavelet transform was conducted on each spectrum, producing an associated scalogram with wavelet transform modulus maxima lines arising from spectral features. By using the ’straightness’ of these lines, it was shown to be possible to identify the presence of ¹³⁷Cs with 95% accuracy, with a number of false positive identifications. Constraining certain metrics within the method reduced the number of false positives at the expense of overall detection accuracy. UK Ministry of Defence © Crown Owned Copyright 2026/AWE

Abstract

This study proposes a 1D Convolutional Neural Network (CNN) architecture designed for real-time particle hit detection, moving beyond traditional rigid triggering thresholds. By processing continuous data streams, including high-noise environments, our model demonstrates superior sensitivity in low Signal-to-Noise Ratio (SNR) regimes and pile-up events. We demonstrate that this approach successfully identifies particle signals and provides precise temporal metadata in real time, offering a robust alternative to conventional signal processing methods. Performance was evaluated on a scintillating detector.

Abstract

Environmental radiation monitoring systems require detectors that can measure radiation in real time and rapidly identify radionuclides. This study developed an artificial intelligence–based spectrum Correction algorithm that converts gamma-ray spectra acquired from polyvinyl toluene (PVT) detectors with limited energy resolution into spectra comparable to those of NaI(Tl) detectors. The proposed model used an artificial neural network (ANN)-based multi-layer perceptron (MLP) regressor with two hidden layers, each containing 64 nodes. Rectified Linear Unit (ReLU) and Adaptive Moment Estimation (Adam) were employed as the activation function and optimizer, respectively. The model was trained using PVT spectra as input and NaI(Tl) spectra as target output data to learn the spectral relationship between the two detector systems. Training datasets were generated using Geant4 Application for Tomographic Emission (GATE) for four radionuclides: Co-57, Ba-133, Cs-137, and Co-60. The simulation results demonstrated that the proposed model successfully generated NaI(Tl) photopeak features from PVT spectra. Representative results for Cs-137 yielded a Pearson correlation coefficient of 0.9996 and a root mean square error (RMSE) of 0.0038. Additionally, the energy resolutions at the 662 keV photopeak were found to be 5.94% for the actual NaI(Tl) spectrum and 5.96% for the predicted spectrum, indicating that the two spectra have highly consistent energy resolution characteristics. These results imply that the proposed model can generate spectra with characteristics comparable to those of NaI(Tl) detectors while maintaining the practical benefits of PVT detectors for radiation monitoring applications.

Abstract

Industrial X-ray imaging for nondestructive inspection requires high spatial resolution and high sensitivity, but phosphor-based detectors face a trade-off because increased phosphor thickness improves X-ray absorption while degrading spatial resolution through optical spreading. This study investigates a measurement-guided Fourier stochastic differential equation for detector X-ray image restoration. The proposed method defines a degradation path in the Fourier domain using measured detector characteristics. The relative modulation transfer function (MTF) models deterministic resolution loss, and the measured noise power spectrum guides stochastic perturbation. A neural restoration network is trained to reconstruct high-resolution images from intermediate degraded states conditioned on low-resolution observations. Experiments on printed circuit board X-ray images compared the proposed method with Wiener filtering, a regression baseline, and a conditional diffusion model (CDM). The proposed method improved low-resolution images and achieved image domain restoration quality comparable to the regression baseline and better than Wiener filtering, while CDM showed slightly lower structural fidelity in representative regions. Frequency domain analysis using MTF, contrast transfer function (CTF), and NNPS showed that both CDM and the proposed method closely followed the high-resolution MTF and that the proposed method showed higher CTF recovery in the mid-to-high frequency range. These results indicate that X-ray image restoration methods should be assessed using both image domain metrics and frequency domain characteristics.

Abstract

Radionuclide identification in high‑purity germanium (HPGe) gamma‑ray spectra can be a challenging and time‑consuming task, particularly when complex samples contain many radionuclides that produce overlapping spectral features. In addition to sample emissions, HPGe detectors measure contributions from naturally occurring and environmental radionuclides, referred to as background. The measured sample spectrum is therefore the sum of sample and background contributions, and in most applications it is desirable to identify only the radionuclides present in the sample itself. While background subtraction is commonly performed using a separate background measurement, accurate identification near critical detection limits remains difficult, especially for low‑count or cluttered spectra.

We present a Siamese Neural Network (SNN)–based approach for radionuclide identification with background subtraction in HPGe gamma‑ray spectra that is generalizable to many detector models and geometries. By directly learning a similarity metric between foreground and background spectra, the method enhances sensitivity to weak spectral features while maintaining low false‑positive rates.

The network is trained on a large dataset containing spectra from over 500 radionuclides, generated using Monte Carlo simulations spanning more than 50 detector and geometry combinations and augmented with measured background data. We present the data generation methodology, training procedure, and performance results on both simulated and measured spectra. The results demonstrate that the SNN approach achieves performance comparable to or exceeding that of traditional spectral analysis techniques. This work provides a foundation for robust, automated radionuclide identification across diverse measurement conditions.

Abstract

Accurate classification of radioactive waste is critical for determining disposal costs and ensuring safety during nuclear power plant decommissioning. However, generic models with simplified geometries and material compositions can lead to systematic misclassification. In this study, a high-fidelity 3D MCNP model of a pressurized water reactor (PWR) was developed to evaluate activation source terms and classify waste based on IAEA Safety Standards Series. The model includes the reactor pressure vessel (RPV), reactor vessel internals (RVI), and bioshield concrete, enabling realistic representation of complex structures. Activation analysis shows that Fe-55, Co-60, and Mn-54 are the dominant radionuclides in both the RPV and bioshield, qualitatively aligning with widely recognized reactor activation data and thereby validating the developed model. Based on specific activity, Fe-55 accounts for over 99% in both structures. However, applying radionuclide-specific activity limits results in a dramatic shift in the dominant contributor. In the bioshield concrete, the contribution of Fe-55 decreases from over 99% to approximately 16%, while C-14 becomes dominant at approximately 76%. This indicates that classification based solely on specific activity can lead to fundamental misclassification. The RPV is classified as low-level waste (LLW), whereas the bioshield meets the criteria for very low-level waste (VLLW). In contrast, generic models fail to accurately represent the reactor configuration, which can result in over-classifying the bioshield as LLW. The high-fidelity approach provides a quantitative basis for reclassification, directly contributing to the reduction of disposal volume. These results demonstrate that high-fidelity activation modeling is essential not only for improving classification reliability but also for avoiding systematic overestimation of waste classes and enabling more economically efficient decommissioning strategies.

Abstract

Nuclear glovebox monitoring must recognise not only which objects are present, but whether their position and orientation indicate an operationally relevant state, such as an inverted container, a handled tool, or an object moving close to a glove during potentially hazardous work. This paper evaluates a two-camera six-degree-of-freedom vision pipeline for advisory operator support in a tabletop analogue of a nuclear glovebox containing a can, pliers, and screwdriver. Offline colour-and-depth pseudo-labels are generated and filtered to train a lightweight object detector and keypoint regressor. During live operation, the system performs object detection, keypoint-based pose recovery, masked depth refinement, temporal smoothing, and safety-state logging. On the validation split, the detector achieved 99.3% mean average precision at an intersection-over-union threshold of 0.5, and the live keypoint-and-depth-refinement pipeline ran at 8.5 frames per second on constrained hardware. These results show that pose-aware monitoring can support prototype glovebox alarms and auditable event records without running a high- capacity pose model online, suggesting a practical route toward advisory systems that improve situational awareness and post- task review while preserving operator judgement. Before such systems could be used operationally, further work is required to establish independent ground truth, model articulated tools, validate longer-term reliability, and build the necessary safety case.

Abstract

Spatially fractionated radiation therapy using tungsten matrix collimators enables high peak-to-valley dose ratios and improved normal tissue sparing. At photon energies near 25 MeV, photonuclear interactions in tungsten induce activation and produce secondary neutrons, potentially affecting both dosimetric performance and radiation safety.

In this work, these processes were investigated using FLUKA Monte Carlo simulations of a modular tungsten-matrix collimator irradiated by a 25-MeV photon beam. Secondary neutron fluence distributions, energy spectra, and absorbed-dose maps in a silicon detector plane were evaluated and normalized to a photon fluence of approximately 1 × 10⁹ cm⁻². The calculated absorbed-dose profile exhibits a pronounced mini-beam structure with a peak dose of 6.16 × 10⁻⁴ Gy, a valley dose of 2.53 × 10⁻⁵ Gy, and a peak-to-valley dose ratio of 24.4. Neutron production is highest at the entrance to the collimator and decreases significantly beyond approximately 5 cm along the beam path, indicating substantial attenuation within the 9 cm tungsten structure. Folding the neutron spectra with energy-dependent fluence-to-dose conversion coefficients yields an ambient dose equivalent of 102 µSv in the collimator region and 0.62 µSv in the phantom region for the same irradiation.

The results demonstrate that photoneutron generation is strongly influenced by the collimator configuration and exhibits distinct spatial and spectral characteristics. Photoneutrons contribute to the dose outside the primary irradiation field, thereby increasing the valley dose and affecting the achievable peak-to-valley dose ratio. These findings underscore the importance of accounting for photonuclear effects during collimator optimization and shielding design for high-energy photon mini-beam therapy systems.

Abstract

Metallic foreign particles are critical safety hazards in lithium-ion battery manufacturing. To address the difficulty of detecting metallic microparticles in single-layer LiFePO₄ cells, this work proposes a nondestructive detection method based on low-energy X-ray transmission imaging. Cu, Fe, and Ni particles are analyzed through attenuation modeling and simulation to evaluate the effects of photon energy, particle material, and particle thickness on transmission contrast. Preliminary imaging experiments are performed using 40keV X-rays, and weak abnormal signals are extracted by normal-sample normalization, filtering enhancement, threshold segmentation, and local SNR analysis. The results show that metallic microparticles can produce detectable local attenuation anomalies in low-energy X-ray images.

Abstract

Current radiomics approaches primarily focus on individual organs or lesions, which limits the ability to describe interactions within the body holistically. To address this limitation, we introduce a graph-based framework for modeling radiomics relationships between organs and lesions in theranostic SPECT/CT imaging. A cohort of 57 patients treated with ¹⁷⁷Lu-based radiopharmaceuticals was analyzed using post-therapy SPECT/CT. Organs were segmented using a deep learning–based approach, and lesions were delineated semi-automatically on SPECT images. From each region, 107 standardized radiomic features were extracted. Interactions between organs and lesions were quantified using a combination of linear similarity measures (Pearson, Spearman, cosine) and a nonlinear dependency metric based on normalized mutual information. These relationships were then represented as patient-specific graphs to capture system-level radiomics organization. The results reveal a non-uniform interaction pattern, where only a subset of organs exhibits strong associations with lesion characteristics. The highest coupling was observed in physiologically active regions such as the kidneys, liver, and spleen, with normalized mutual information values reaching approximately 0.66. Across all regions, nonlinear dependencies consistently showed higher values than linear metrics, indicating that complex relationships dominate radiomics interactions. Network representations further demonstrated a densely connected structure without prominent hubs, suggesting a distributed interaction model. Stability analysis confirmed reproducible patterns across patients, with near-perfect rank consistency and high agreement among top-ranked regions. These findings demonstrate that radiomics interactions between organs and lesions are structured, nonlinear, and reproducible. The proposed framework enables system-level interpretation of imaging data and may support the development of more informative biomarkers.

Abstract

In non-destructive testing (NDT), simulation is used to study radiation–matter interactions for various cases. CIVA software includes a radiographic testing (RT) module, which combines analytical and Monte Carlo (MC) approaches. The analytical component models primary photons, while the MC component handles scattered photons — a process that can become computationally intensive in complex configurations. Building on this, we introduce an estimator based on a split exponential track length method (seTLE) in the MC component of CIVA-RT, designed to reduce statistical variance and computational cost. This next-event estimator increases the probability of photon contributions to the detector by creating multiple pseudo-photons during the simulation. Its performance is evaluated using a convergence index and a Figure of Merit. Three configurations are studied: a simple configuration, a step wedge configuration with a distant detector, and an industrial weld case. Results show that the estimator provides significant efficiency gains. A detailed analysis of seTLE simulation shows that rare high-weight photons can cause sudden signal variations. These results demonstrate that the approach provides efficiency gains, making high fidelity simulations more tractable in industrial contexts.

Abstract

The CEA develops and uses airborne detectors to carry out gamma mapping missions for various private and public entities, and various national and international authorities. The main objectives of these missions are : site monitoring (natural and anthropic contributions to environmental radioactivity), mapping contaminated areas, detecting and identifying radionuclides and radioactive sources.

Methods for analyzing the spectra acquired during these missions are being studied, with a view to improve decision thresholds and detection limits for various radionuclides of interest. A spectral analysis method called "spectral unmixing" has been developed, using a full spectrum analysis instead of the reference method based on regions of interest. This algorithm takes into account the response of the detector under study to an individual radionuclide, and estimates the proportions of the different radionuclides using a likelihood estimator. This method was shown to be accurate for reconstructing radionuclide activities, with improved sensitivity compared to the reference method.

The application of this algorithm to airborne detectors raises a number of challenging problems, such as taking into account the variability of spectral signatures linked to the environment and source shielding. An algorithm has been developed to address this problem. The algorithm was validated on simulated and experimental data, and improved the results of an airborne measurements campaign. This method can be implemented on the measurement system for real-time radionuclide identification.

Abstract

Scatter correction is a vital tool in enhancing SNR and contrast recovery in PET; however, existing techniques are computationally costly, need priors, and fail at lower counts. Our focus is to remedy these concerns using a novel closed-form solution. The technique differentiates itself by considering multi-Compton scatter (>2 scatter events) that twice convolve the single-scatter energy distribution revealing 9 states of joint probability density function. This then forms the Gram covariance matrix in our L2-regularized regression with Bhattacharrya weighting: an index of the probability density function overlap establishing appropriate penalties. Lastly, a quadratic cost function is implemented, and the solution is determined via gradient optimization method. Preliminary results from a Utah Phantom mimic’s coincidence data simulated in our GenII PETCoil in GATE v.8.2 determined a 10-fold decrease in magnitude of the L2 lambda compared to our previous technique suggesting better conditioning. The global scatter fraction for the phantom, compared to the ideal estimate from GATE came remarkably close: ~0.08% with the median error per sinogram bin centered around 0.00% (99.7% CI [0.00%,0.00%]) within GATE's ideal scatter fraction within an average of 360-410 counts/sinogram bin after correction (using SF = 1-True Fraction) suggesting a near perfect scatter correction. To determine the potential benefit of a prior joint probability density function of the anatomical Utah Phantom, we reran the algorithm receiving a global scatter fraction error of 3.13% with a per sinogram bin error of -1.12% (99.7% CI [-1.17%,-1.06%]) indicating an overcorrection. Our results conclude the following: the KN derivation closed-form solution approaches the ideal and a probabilistic prior is not needed.

Abstract

This study proposes a hybrid semantic–keyword analysis framework for automatically identifying and classifying nuclear fuel cycle-related activities using open-source text data. In the safeguards system of the International Atomic Energy Agency, continuous monitoring and analysis of nuclear fuel cycle activities are required; however, existing approaches rely heavily on expert-driven manual review, limiting efficient processing of large volumes of unstructured text data. To address this issue, nuclear fuel cycle-related news articles collected from Google News were used to construct profiles for seven nuclear fuel cycle activities. Each activity was represented by combining semantic information derived from Sentence-BERT embeddings with keyword importance obtained using class-based TF-IDF. Subsequently, key sentences were extracted from IAEA news articles, and activity classification was performed by integrating semantic and keyword similarity for each activity category. Performance evaluation results showed that the proposed method achieved a Top-1 accuracy of 57.14%, a Top-3 accuracy of 85.71%, and a mean reciprocal rank of 0.738. In addition, the actual activity category was generally included within the top candidate groups, showing that the proposed framework reflects the semantic relationships and complex characteristics among nuclear fuel cycle activities. This study demonstrates the feasibility of automating open-source-based nuclear fuel cycle activity analysis and suggests the potential for future extension to an Additional Protocol-based automatic declaration target identification system. The results also highlight the applicability of data-driven analytical approaches for safeguards and regulatory support.

Abstract

High-performance industrial gamma imaging under count-limited conditions is challenging, particularly for medium- and high-energy radionuclides where photopeak statistics are limited. However, some non-photopeak Compton-related events still retain directional information. We propose a directionally informed multi-energy-window reconstruction method that selectively incorporates such events. Monte Carlo spectral decomposition was used to define energy windows and estimate useful-event fractions. Window-specific projection data were paired with matched system matrices and jointly reconstructed using weighted MLEM.

The method was evaluated with a self-collimating industrial gamma camera for ¹³⁷Cs imaging. In point-source experiments, 30 s multi-window reconstruction achieved localization accuracy comparable to 90 s photopeak-window-only MLEM. In structured-source and short-frame dynamic imaging, it reduced background artifacts and improved localization stability. These results show that selective use of directionally informative non-photopeak events can improve event utilization and reconstruction robustness under low-count industrial gamma imaging conditions.

Abstract

Three-dimensional reconstruction of fusion neutron sources is critical for diagnosing burn-region morphology and implosion asymmetry in inertial confinement fusion, yet the recoverable morphology is severely limited by sparse and geometrically constrained diagnostic views. This work proposes an optical surrogate framework to study information-driven sparse-channel layout design for fusion neutron source reconstruction, offering controllable source morphology, low cost, and ease of iteration compared to direct neutron diagnostics. The core motivation is to treat limited-view reconstruction not only as an inverse-algorithm problem, but also as a constrained layout-design problem. A conceptual design employing Latin hypercube sampling generates candidate layouts within admissible geometries, followed by greedy selection. The optical surrogate tests the shared inverse-problem structure using movable point-like optical sources, directional fiber channels, rather than reproducing neutron transport physics. Preliminary fiber-response measurements reveal finite-acceptance soft-channel behavior with a transverse full width at half maximum of approximately 2.7 mm. Theoretical reconstruction using a discrete multi-point-source design yields a detector-domain residual of 0.038, a source-correlation coefficient of 0.998, and a maximum relative strength-ratio error of approximately 7.2%. These results establish a forward-model basis for subsequent layout optimization and future migration to neutron-response matrices.

Abstract

Industrial X-ray computed tomography (CT) is widely used for non-destructive inspection and generation of geometric models. However, conventional voxel-based CT reconstruction on a fixed Cartesian grid causes partial-volume blurring and possible geometric errors when converted into surface or CAD models. This work presents an interface-adaptive tetrahedral mesh-based iterative reconstruction framework for industrial CT. The method estimates nodal attenuation coefficients directly on an unstructured mesh. X-ray projection and backprojection are performed by an NVIDIA OptiX-accelerated projector, which computes ray-mesh intersections, segment lengths, and barycentric weights, and backprojects residuals using the SIRT algorithm. To improve boundary representation, the reconstructed field is used to detect interface-related edges, locally refine the mesh, and displace vertices. For each candidate vertex, a local attenuation-gradient direction is estimated from neighboring nodes. The neighbors are divided across the estimated normal, and the vertex is shifted toward the attenuation mid-level between regions while enforcing element quality constraints. Simulated projection studies showed the proposed adaptive mesh reconstruction preserved overall structure while producing sharper material boundaries than voxel-based SIRT. Global structural similarity improved, particularly in regions containing material interfaces. These results indicate the framework reduces boundary blurring, supporting more accurate defect sizing and dimensional measurement in industrial testing.

Abstract

This study presents a forward simulation of high-energy X-ray phase-contrast imaging (XPCI) for semiconductor package inspection using an X-ray microscopy (XRM)-derived structural phantom. A projected phantom from an experimental laboratory XRM reconstruction stack was used as a geometry prior rather than quantitative material ground truth. The projected grayscale phantom was percentile-normalized and mapped to simplified Cu/Si refractive-index distributions. Under the proposed long-beamline high-energy synchrotron XPCI conditions, the simulated XPCI intensity and retrieved phase images reproduced characteristic package structures, including stacked layers and metal interconnect regions. The finite-source-induced blur was on the order of the detector pixel size, while the selected intra-bump feature showed higher local visibility in the retrieved phase image than in the simulated intensity image. This work extends idealized-phantom simulations to a laboratory-XRM-reconstructed semiconductor package structure under long-beamline high-energy synchrotron XPCI conditions.

Abstract

X-ray scannings are widely adopted by customs to control containers. Transmission and backscatter X-ray imaging are often jointly employed to enhance material discrimination and detection robustness in cargo inspection systems. However, the two modalities typically operate at substantially different photon energy levels, leading to energy-dependent interaction volumes and effective source geometries that cannot be simultaneously aligned. As a result, image registration becomes a prerequisite for joint analysis and decision-making based on transmitted and backscattered measurements, yet it is fundamentally ill-posed because consistent spatial correspondence between the two imaging processes is not directly observable. In this work, we investigate this identifiability issue in joint transmission--backscatter X-ray image registration by explicitly modeling energy-dependent source geometry and imaging physics. We show that mismatched photon energies can induce structural ambiguity, such that multiple spatial alignments may remain similarly plausible under joint interpretation.

Abstract

To facilitate the highly efficient and accurate surveying of radioactively contaminated areas, this study presents the development of an advanced UAV-mounted integrated survey system capable of the simultaneous, real-time measurement of 3D geospatial information and environmental radiation data. The hardware architecture is robustly constructed, integrating two sets of two-inch LaBr3(Ce) detectors for high-resolution gamma spectrometry. This is combined with a LiDAR sensor for acquiring precise 3D terrain topography, a GPS, and a laser altimeter to ensure accurate spatial referencing. To guarantee operational stability during flight, the payload design incorporates specialized vibration-damping mounts and a lightweight 1.5-millimeter aluminum housing. Furthermore, a high-speed LTE router network was introduced to reliably handle the transmission of large-capacity data streams to the ground station. 
The practical efficacy of the system was rigorously validated through extensive flight tests conducted over radioactively contaminated regions in Fukushima. When aerial measurements were mathematically converted to dose rates at one meter above ground level, the results demonstrated highly satisfactory performance, showing a strong statistical correlation with baseline data obtained using conventional backpack-type ground survey equipment. Crucially, by mapping real-time energy spectra directly onto the acquired 3D geospatial models, the system successfully enabled the comprehensive, multi-dimensional identification of specific high-level contamination hotspots. Moving forward, by integrating AI to further refine advanced topographical correction algorithms, this autonomous system is expected to significantly enhance the rapid and precise generation of contamination maps, providing critical support for emergency response.

Abstract

To meet the on-site inspection requirements for uranium materials, a handheld uranium enrichment analyzer based on a LaBr3 detector and the relative detection efficiency self-calibration method was developed. The adopted LaBr3:Ce crystal has a size of 38.1mm in diameter and 25.4mm in thickness, with an energy resolution of 2.9% at 662keV. The handheld uranium enrichment analyzer is equipped with a 4.5-inch touchscreen. It features overall dimensions of 145×75×58mm and a weight of 800g, supporting continuous operation for more than 8 hours. Test results from six NBS standard samples show that for uranium samples with a mass≥200g, a ²³⁵U enrichment≥0.711%, and a measurement time≥1800s, the relative deviation of the measured ²³⁵U enrichment is less than 8.0%. This device can be applied to the verification and measurement of ²³⁵U enrichment in uranium materials.

Abstract

Boron Neutron Capture Therapy (BNCT) and high-flux pulsed neutron sources such as J-PARC require reliable real-time neutron monitoring with minimal perturbation of the irradiation field. We have been developing miniature optical-fiber-based neutron detectors using small Li-glass scintillators mounted at the fiber tip. Conventionally, the scintillator pieces were prepared by crushing bulk Li-glass, producing irregularly shaped fragments that introduce geometric uncertainties and complicate quantitative response evaluation by Monte Carlo simulation. To overcome this, an aerodynamic polishing technique was established to shape Li-glass fragments into well-defined spherical particles with diameters of 100 to 500 μm without significant degradation in light yield. The spherical geometry enabled quantitative prediction of the wall effect through PHITS Monte Carlo simulations with a quenching model based on Birks’ law, and the simulated pulse-height spectra showed excellent agreement with experimental data obtained from spheres of various diameters (220–450 μm). This approach provides a robust framework for the precise design and manufacturing of miniature neutron detectors for high-intensity fields.

Abstract

The Coded Cube Camera for Gamma-ray (C3G) is a 4π gamma-ray imager designed to estimate gamma-ray source distributions from direction-dependent detector response patterns.  In this study, we developed a 24-detector C3G consisting of 24 HR-GAGG scintillators and applied it to a plant radioisotope (RI) imaging experiment using ¹¹CO₂ gas.  In the plant RI imaging experiment, approximately 300 MBq of ¹¹CO₂ gas was supplied to a plant placed in an experimental chamber to image the transport of ¹¹C-labeled photosynthates within the plant.  During this experiment, the C3G was placed outside the chamber, and the dynamics of ¹¹CO₂ gas were continuously measured with an acquisition time of 1 s per measurement.  As a result, the inflow of the gas into the chamber was clearly visualized.  In addition, the reconstructed image intensity decreased exponentially over time.  The half-life estimated from this intensity change was 20.70 min, which was in good agreement with the known half-life of ¹¹C, 20.34 min.  These results demonstrate that the 24-detector C3G can visualize dynamic gaseous radioactive sources and estimate temporal changes in radioactivity from reconstructed image intensity.  This capability is expected to be useful for applications such as radioactive gas leakage monitoring, environmental monitoring, and nuclear security.

Abstract

Characterization of nuclear fuel in fuel debris is essential for the decommissioning of the Fukushima Daiichi Nuclear Power Station. The fast fission neutron coincidence counting (FFCC) method, which is less sensitive to neutron absorbers than conventional thermal-neutron-based techniques, was investigated using Monte Carlo simulations with PHITS. A cylindrical unit can containing fuel debris was modeled, and four detector configurations with 24–72 liquid scintillator detectors were compared. Fast neutron coincidence counting efficiency decreased as the number of detectors decreased, whereas the quantification limit did not vary monotonically; the 56-detector configuration provided the best performance. This indicates that an optimal number of detectors exists, as gains in coincidence events are offset by reduced efficiency in distant detectors and increased background counts. Evaluation of the effective dose showed that shielding with lead and boron-loaded polyethylene reduced the dose around the system below the access limit for personnel, without significantly affecting measurement performance. These findings provide useful guidelines for the design of practical FFCC systems.

Abstract

The ASP neutron facility at AWE – Nuclear Security Technologies (AWE-NST) Aldermaston, is undergoing upgrades to both the accelerator beamline and the supporting diagnostics.  This is driven by evolving customer requirements, particularly its increasing use as a platform for fundamental nuclear data measurements. Neutron flux monitoring on ASP is provided by fission chambers and is supported by external activation foil calibrations and internal cross-calibration measurements. A recommissioned associated-particle technique provides improved accuracy for lower-fluence measurements, while ongoing work is improving neutron spectral characterisation using a range of techniques including activation, Single crystal chemical vapour deposited diamond detectors, proton recoil and Helium - 3 spectrometers. These activities reduce uncertainties in current measurements, and establish the diagnostic techniques required for a future facility (VENTURE) capable of delivering higher fluxes and a wider range of quasi-monoenergetic neutron energies. The calibration and diagnostics strategy for the VENTURE facility is outlined, with an emphasis on delivering the well characterised source required for nuclear data experiments.

UK Ministry of Defence © Crown owned copyright 2026/AWE

Abstract

As the demand for the detection of special nuclear materials increases, the use of imaging instruments that can sensitively image both gamma-ray and neutron signatures has become necessary. The radioactive waste streams can contain neutron-emitting  radioactive materials, hence localisation of neutron sources plays an important part in nuclear decommissioning. Neutron detector design based on the MURA revealed promising results when tested in security applications. Due to the requirement of ensuring high detection efficiency for both neutrons and gamma rays, the plastic scintillator used is typically thick. As demonstrated in our previous study, the dual-layer design exhibits better imaging performance and produces fewer artifacts. In this paper, we propose a dual-layer DOI (depth of interaction) detector design and have successfully developed a prototype module. The first and second layer of the crystals are clearly resolved from the flood histograms, demonstrating its capability to provide two-layer depth encoding. The energy resolution of each layer is improved by 50% compared with that of a single-layer detector with the same thickness, and the neutron/gamma discrimination figure of merit (FOM) is also significantly enhanced. Validation and testing of the imaging performance are currently underway.

Abstract

Scintillator-based thermal neutron detectors are utilized at scattering beamline facilities around the world. The Anger camera is one example that was designed and deployed at Oak Ridge National Laboratory. The camera uses enriched ⁶Li-glass (GS20) that offers advantages over the widely used ³He gas proportional counters, including sub-millimeter spatial resolution, higher count rate capability, and high neutron absorption efficiency. However, GS20 has higher γ-ray sensitivity due to its solid state, potentially reducing the neutron signal to background ratio. Gamma rejection using pulse height discrimination is feasible, but it suffers when there is greater overlapping between the neutron and gamma spectral features in intense background fields, like with certain beamline configurations. Additionally, pulse shape discrimination is unfeasible with GS20. Previous work demonstrated a new approach to discrimination by optimizing the scintillator shape and reducing its size. In this work, we demonstrate improved gamma discrimination by segmenting the GS20 scintillator layer of the Anger camera. The layer is divided into several thinner layers with the same overall thickness, and coupled together using non-scintillating quartz layers. The neutron absorption products fully deposit their energy in the GS20 layers due to their significantly shorter range compared to the Compton-scattered electrons from γ-rays that only partially deposit their energy. This allows for better separation between the respective features on the pulse height spectrum, and therefore improved discrimination, while maintaining the same high neutron absorption efficiency.

Abstract

This study demonstrates organic scintillator-based neutron coincidence counting as a promising approach for photonuclear nondestructive assay (NDA) of nuclear materials. Four organic scintillators—organic glass scintillator (OGS), EJ-276D, stilbene, and BC-501A—were systematically evaluated in terms of fast-neutron detection efficiency, pulse shape discrimination (PSD) capability, and timing response. In addition, three-detector coincidence counting systems based on OGS and stilbene detectors were constructed and tested using a ²⁵²Cf source. Stilbene and BC-501A exhibited excellent PSD performance, EJ-276D showed high fast-neutron detection efficiency, and OGS provided superior timing response. The OGS-based system successfully measured neutron double- and triple-coincidence events with and without lead shielding, whereas the stilbene-based system showed a marked reduction in coincidence events under unshielded conditions. These results indicate that combining PSD, neutron coincidence counting, and gamma-ray shielding enhances the extraction of correlated neutron signals in photonuclear NDA. The proposed OGS-based approach is a promising for active interrogation systems in nuclear security and safeguards.

Abstract

CEA Cadarache currently manages over 6,000 historical 870 L nuclear waste packages. To ensure long-term safety, French regulatory authorities require a precise characterization of each unit, specifically regarding structural integrity, fissile material mass and total activity of alpha emitters. However, the high density and volume of 870 L concrete waste packages result in significant signal attenuation, making standard passive gamma spectrometry or passive and active neutron measurements inefficient to assess the mass of nuclear materials. To maximize the signal-to-noise ratio, a specific Active Photon Interrogation (API) approach is developed inside the CINPHONIE irradiation bunker of CHICADE facility, at CEA IRESNE, Cadarache, France. This method makes use of inorganic BGO scintillators to measure high-energy delayed gamma rays from short-lived photofission products induced with a high-energy bremsstrahlung photon source, i.e., a 18 MeV LINAC. A Bayesian framework is applied to infer both the mass and the position of nuclear materials from angular acquisitions. This article is focused on testing API framework in the case where the source of the signal is located in a homogeneous environment. Based on the promising results presented here, future developments will focus on the heterogenous case.

Abstract

Radiation monitoring systems (RMS) are essential for continuous radiation surveillance and rapid alarm generation in nuclear facilities. This paper presents the development and performance evaluation of a miniaturized Local Processing and Display Unit (LPDU)-based RMS designed to reduce installation space and radioactive waste generation while maintaining reliable monitoring performance. The developed RMS adopts a 1:4 architecture that enables a single LPDU to process and display signals from four independent detectors. The system was implemented using a Texas Instruments TMS2800137-based edge device and a Canberra G64IC detector. Firmware was developed to support pulse counting, EEPROM operation, RS-232/RS-485 communication, and real-time data processing. A dedicated communication protocol with Cyclic Redundancy Check (CRC) verification was also implemented to enhance data integrity. In addition, a Centered Significant Test (CST)-based adaptive filtering algorithm was embedded in the edge device to improve noise suppression and detector signal monitoring under field conditions. Performance verification tests were conducted in accordance with IEC 60532:2010 using a calibrated Cs-137 gamma radiation source. The developed RMS satisfied the acceptance criteria for linearity, statistical fluctuations, response time, and alarm response time and stability. The maximum linearity deviation was 2.1%, the coefficient of variation was 0.0203%, and all response times were below 6 s. These results demonstrate that the proposed RMS provides reliable radiation monitoring performance in a compact multi-detector configuration.

Abstract

Silicon photomultiplier (SiPM) based thermal neutron detectors have been developed at Oak Ridge National Laboratory for single crystal diffraction instruments. They have potential as good replacements for current 3He detectors on neutron reflectometers due to their high spatial resolution and fast scintillator decay time. The current software-implemented position fitting on these detectors is limited due to the processing required for the recursive least-squares algorithm. In this work, we present a fast method for determining neutron position that can be run on the Field-Programmable Gate Array (FPGA) chip that is already installed on the detector. This method compares the collected event data with pre-determined shapes stored on FPGA memory in parallel, and determines the one with the least linear error. Using this method for neutron positioning takes less time than the hardware defined dead time. Count-rate and spatial resolution results collected at the High-Flux Isotope Reactor are presented, with output count rates > 500 kcps achieved, an order of magnitude greater than the software fitting method.

Abstract

Analysis of lithium ore grade is industrially significant for efficient lithium resource development and utilization, enabling improved production efficiency and reduced costs. Lithium has a large neutron absorption cross-section, and changes in lithium grade can affect the neutron lifetime, which is inversely proportional to the macroscopic absorption cross-section of ore. In this paper, a method for determining lithium grade through neutron lifetime measurements based on an e-LINAC driven photoneutron source is applied to the online analysis of lithium ore grade. Since lithium does not emit gamma rays upon neutron absorption, the neutron lifetime is measured from the time distribution of capture gamma rays induced by neutrons in the matrix elements of the medium. The feasibility of this approach has been demonstrated in experiments using various lithium nitrate solutions as simplified media. Meanwhile, the effect of neutron diffusion was investigated by simulations. The potential of neutron-induced prompt gamma analysis for reducing interference caused by variations in matrix composition is discussed. Further simulation studies and system design are in progress.

Abstract

This work presents an EJ309 organic liquid scintillator system for neutron–photon detection in mixed radiation fields. The performances achieved are illustrated through two complementary use cases. The first concern photoneutron production in pulsed high-energy photon fields generated by a TrueBeam medical electron LINAC and it has been used to characterise neutron–photon pulse-shape discrimination, integrated-charge observables, pile-up and saturation signatures, and the response of the acquisition chain under short intense irradiation pulses. This configuration simulates closely enough the energies involved in high-altitude cosmic-ray neutron production allowing the use of the previously qualified detection system to be used for neutron monitoring at the Jungfraujoch research station (Switzerland, at 3600 m), where the neutron field is mainly produced by atmospheric secondary cascades, including proton- and alphainduced spallation reactions. The detection set-up and analysis framework will be used to measure time-resolved neutron-like and photon-like event populations, pulse-shape discrimination distributions, and spectral observables over long acquisition periods. The work demonstrates the use of EJ309 scintillators for fast-neutron detection in controlled photoneutron fields and supports their application to natural high-altitude cosmic-ray environments, atmospheric radiation monitoring, and photoneutron studies relevant to intense high-energy photon fields.

Abstract

The point kinetics model can be extended to multiple regions or energy groups to account for system heterogeneities. This paper presents an extension of the Feynman-α technique in diverse measurement configurations relevant to nuclear safeguards. This extension is achieved by incorporating phase-type distributions into the heuristic derivation of the Feynman model. Phase-type distributions effectively capture the absorption times of finite-state Markov processes. The resulting expressions are specifically tailored to each experimental configuration. In a bare source configuration using an array of ³He-filled proportional counters, the neutron decay exhibits bimodal behavior. When a supplementary reflector is added to the setup, a third term emerges in the expression, as expected. However, the shape of this additional term depends on the type of reflector, due to assumptions in the system’s transition matrix. When fitted to simulated datasets, these new models show the fundamental decay mode of the detector (approximately 70 µs) and are validated with a 95% confidence χ2 test.

Abstract

A two-stage Compton camera prototype was developed for compact gamma-ray imaging applications in medicine [1]. Each camera layer consists of four monolithic 14 × 14 × 25.4 mm³ LaBr₃:Ce scintillation detector assemblies manufactured by CapeScint, each coupled to a single-channel silicon photomultiplier (SiPM) readout. This configuration does not provide intrinsic position sensitivity within individual crystals.

Detector performance was experimentally characterised and modelled using the TOPAS Monte Carlo toolkit, with validation performed against standard gamma-ray calibration sources [2]. Data acquisition was performed using a XIA PIXIE-16 fully digital signal processing system operating in list-mode, enabling event-by-event energy and timing extraction with digital pulse processing [3]. The detectors achieved an energy resolution of approximately 3.4% at 662 keV and a coincidence timing resolution of 988(17) ps.

Imaging performance was evaluated using a GBq ¹³⁷Cs source at multiple source-to-detector distances. Reconstructed experimental images were compared with validated simulations, with simulations yielding a full width at half maximum (FWHM) of 3.36(1) mm and an angular resolution measure (ARM) of 1.47(13)^o.

Results demonstrate agreement between experiment and simulation, validating the modelling framework and supporting the feasibility of compact LaBr₃:Ce-based Compton camera systems for medical gamma-ray imaging applications.

Abstract

Mixed-field irradiation facilities require accurate and traceable dosimetry to support the qualification of electronic components and systems for accelerator environments. This work presents an updated dosimetry and beam-condition characterization of the CERN High-energy AcceleRator Mixed-field facility (CHARM), focusing on the comparison of different irradiation positions and on the extension of the available test conditions to new high dose-rate and high fluence locations close to the target. The study combines active and passive dosimetry methods, including CERN RadMon radiation monitors, radiophotoluminescent glass dosimeters and alanine dosimeters, to evaluate Total Ionizing Dose and fluence-related quantities under representative operating conditions. Measurements acquired during commissioning are compared with data collected during user operation to assess the stability and reproducibility of the delivered radiation fields. In addition, the effect of upstream IRRAD samples on the residual beam reaching the CHARM target is investigated using the IRRAD beam position monitors and the multi-wire proportional chamber installed at the end of the T8 beam line. Variations in the upstream material configuration are shown to affect the transported beam properties, including divergence, with a measurable impact on the resulting mixed field. The combined dosimetry and beam instrumentation data are used to validate the new irradiation positions, compare independent dosimetry techniques and define operational reference conditions for CHARM tests. These results improve the traceability of dose and fluence delivery in mixed-field accelerator irradiation campaigns and support more robust interpretation of radiation effects tests performed at CHARM.

Abstract

Gamma-ray spectroscopy is a fundamental technique for identifying radioisotopes in safety-critical applications. However, in scenarios such as nuclear accidents, high-temperature and high-radiation environments severely degrade photomultiplier tube performance. Conventional temperature compensation systems face physical voltage limits, and prior deep-learning algorithms fail because they require preserved spectral shapes. To overcome these limitations, this study presents a comprehensive strategy to enhance detector stability. First, an active high-voltage control system mitigates initial temperature-induced gain loss. Because this hardware compensation eventually reaches its physical safety limits, the system integrates a deep learning (DL) technique – bidirectional long short-term memory network – to interpret residual spectral distortions, successfully achieving accurate multi-isotope quantification up to 185 °C. Second, to address radiation-induced damage localized within the photomultiplier tube, a targeted active thermal annealing protocol was applied. Following a 1.55 kGy cumulative gamma irradiation, detector output dropped to 23.5% of its baseline. Subsequent thermal annealing of the photomultiplier tube body at 145 °C successfully restored the output to 89.4%. Ultimately, this integration of hardware, software, and active recovery lays a foundation for robust radionuclide identification under concurrent high-temperature and high-radiation constraints.

Abstract

Large cargo scanning presents a unique challenge for national security. Container size limitations pose challenges to tomographic imaging. This work proposes a new stationary, limited angle tomographic transmission scanner geometry with a unique β/γ source utilizing a pair of fixed flat panel detectors, known as Gamma Tomography (GT). A Gate 10 simulation evaluates the imaging capabilities of the scanner and explores time-of-flight (TOF) and energy discrimination as electronic collimation methods to reduce scattered radiation and improve image quality. A 600 μm thin plastic scintillator sheet encasing 100mCi of ⁶⁰Co as a near detector locally captures nearly all emitted β particles without significant γ interactions, while a far detector panel of 2 cm thick LYSO crystals captures the ~1.25 MeV energy ⁶⁰Co γ photons. Both panels were simulated as 50x80 cm with a thin, 2 cm thick, multi-material test object consisting of water, aluminum, titanium, steel, and uranium in the center of the field of view. The same simulation dataset was reconstructed with/without two separate electronic collimation schemes: valid coincidences with a 100 keV energy window around the ⁶⁰Co photopeaks, and valid coincidences with a 100 ps time resolution. Despite the short acquisition time (18sec), reconstructed images are low noise, with all objects visible above background and distinguishable. The images show attenuation coefficient under-estimation, due to object scatter. TOF and energy discrimination underestimate μ values by 25.8% and 34.3% respectively, but a large dynamic range and very high SNR in μ value is achieved for all reconstructions (SNR-water: 46, SNR-Uranium: 1369). Using time information in lieu of energy allows events scattered in the detector to be used to improve image quality and SNR, but further investigation will include achievable time resolution, panel construction, and source geometry optimization.

Abstract

Alpha particle autoradiography is an emerging modality for mapping the biodistribution of alpha-emitting radionuclides in preclinical studies, supporting the development of radiopharmaceuticals for targeted alpha therapy (TAT). Considerable effort has been directed toward achieving high spatial resolution, with recent advances emphasizing single-event processing and energy-resolved detection to identify the nuclide of origin of individual alpha particles. Our previous work validated a spectrum decomposition method for dual-isotope imaging of ²²⁵Ac and ²¹³Bi using a spectroscopic alpha imaging detector, confirming its feasibility with therapeutic amounts of ²²⁵Ac. However, the active area of that system was constrained to 2.5 cm × 2.5 cm, limiting its applicability to small fields of view. To overcome this constraint and accommodate simultaneous imaging of multiple tissue sections or organs from larger animals, we developed a large-area spectroscopic alpha imaging detector with an active area of 5 cm × 5 cm, representing a fourfold increase over the prior system. The detector achieved an energy resolution of 4.3% FWHM at 5 MeV and a spatial resolution of approximately 600 μm, representing a substantial improvement over the prior system. Future work will include demonstration of dual-isotope spectral imaging of ²²⁵Ac and ²¹³Bi to support preclinical TAT evaluations.

Abstract

Scintillator-on-fiber (SOF) detectors utilizing Liglass scintillators show strong potential for neutron flux monitoring in intense radiation environments due to their high radiation hardness and fast response. Previous studies have typically employed fragmented scintillators in SOF detectors due to the difficulty of μm-scale machining. In this work, Li-glass scintillators with well-defined dimensions (1×1×0.5 to 0.25×0.25×0.5 mm³) were precisely fabricated using laser cutting, enabling a systematic investigation of the effect of scintillator size on neutron response linearity in an acceleratordriven neutron beam. With pulse-shape discrimination (PSD), the results show that neutron signals measured with scintillator sizes larger than 0.5×0.5×0.5 mm³ become blurred in the P_SP -channel scatter plot due to gamma-ray contributions and pulse pile-up at an accelerator current of 250 μA. In contrast, the smallest scintillator (0.25×0.25×0.5 mm³) maintains clear separation between neutron and gamma-ray signals. Analysis of the maximum shift error σ_m of the neutron peak centroid further indicates that this smallest scintillator effectively suppresses pileup, achieving σ_m = 0. In terms of linearity, scintillators with sizes below 0.75×0.75×0.5 mm³ yield coefficients of determination R² > 0.99 when processed using a two-dimensional fiducial cut, satisfying the requirements for neutron flux monitoring in boron neutron capture therapy (BNCT). These results demonstrate the feasibility of optimizing Li-glass scintillator size and volume in SOF detectors to achieve reliable neutron flux monitoring in BNCT applications.

Abstract

Radiation accident response requires real-time source-direction and radionuclide information. However, conventional systems are limited in simultaneously achieving radionuclide identification, direction estimation, 360° monitoring, real-time processing, and multiple-source detection. To address this, a directional radiation monitoring system (DRMS) using the depth-of-interaction (DOI) method was developed previously, and its feasibility for 360° monitoring and multiple-source detection was verified. This study evaluated the real-time direction detection performance of the developed DRMS under moving-source conditions. The DRMS consisted of a CsI(Tl) scintillator (1" diameter, 4" height), two photomultiplier tubes coupled to both ends, a cylindrical collimator with eight slots arranged in different directions, and an FPGA-based DAQ. The system calculated gamma-ray energy and DOI using the sum and ratio of the two PMT signals, respectively, and generated radiation distribution maps in real time based on the slot direction corresponding to the estimated DOI layer. Evaluation was conducted assuming an accident scenario in which a 5 TBq radiation source mounted on a vehicle moved at 40 km/h, 100 m from the system. To reproduce this in a laboratory, measurements used a 3.17 MBq ¹³⁷Cs source moved at 4.23 cm/min, 1 m from the system, under the same count-rate condition. Energy windows were applied to the data, and the nuclide was identified when counts within its window satisfied the criterion. A nuclide distribution map updated at fixed count intervals to reflect the real-time source position and an accumulated distribution map reflecting the total response were generated. The results confirmed that the system could visually reflect the position change and movement trajectory of a moving source. These results demonstrate the potential of the DOI-based DRMS for real-time source-direction monitoring in radiation accident response and mobile radiation detection.

Abstract

Measuring the radioactivity concentrations of airborne dust is crucial for selecting appropriate protective equipment for radiation workers and assessing internal exposure of them. In this study, we aim to develop a portable dust monitor that can measure alpha and beta rays simultaneously without being affected by ambient gamma rays. For the development, we utilize the clothing contamination monitor, which has been successfully put into practical use in previous research.
We investigated whether it was possible to distinguish between alpha and beta rays based on their respective energy spectra by irradiation tests. The results showed that by setting appropriate regions of interest (ROIs), we were able to minimize the influence of gamma rays and successfully distinguish and measure alpha and beta rays.

Abstract

This work systematically characterizes how transistor geometry governs neutron displacement damage (DD)-induced degradation in silicon MOSFETs. A monolithic single-chip array integrating 1.8 V and 3.3 V NMOS/PMOS devices with seven channel widths and 1-, 5-, and 10-finger layouts was fabricated as the device under test, and a hierarchical 1 GHz multiplexing readout was developed to track 84 transistors in real time during irradiation at TRIUMF (10⁶–10¹⁰ n/cm², five incremental steps, annealing monitored up to 40 days). For 1.8 V NMOS devices, the post-irradiation output-voltage shift scaled with channel width, while the threshold-voltage shift after 40-day annealing varied strongly with finger configuration at the same total channel width. These results indicate a strong dependence of DD-induced degradation on finger topology, beyond what channel-width scaling alone can explain, with 10-finger layouts exhibiting anomalous nonlinearity attributable to parasitic coupling and non-uniform damage.

Abstract

This paper presents a solid-state detection technology designed to improve large-area surface contamination measurements and enable integration into contamination and clearance equipment or robotics.

The primary objective is to replace photomultiplier-based detectors with compact architecture using solid-state photosensors and embedded digital processing, while maintaining or improving detection performance.

The proposed detector consists of a 579cm² sensitive surface divided into two independent channels, each coupled to integrated electronics performing biasing, signal processing, self-test, and communication. A daisy-chain configuration was implemented to evaluate system scalability, allowing up to fifty detectors to operate synchronously within a single network. Measurements demonstrate detection homogeneity better than plus or minus fifteen percent over the active area and typical hemispherical efficiencies up to forty-six percent for high-energy beta emitters. The acquisition architecture enables rapid readout compatible with moving platforms, with sampling rates reaching one hundred measurements per second for moderate network sizes.

Compared with legacy systems, the proposed solution significantly reduces mass while improving robustness and insensitivity to magnetic fields and eliminates the need for complex external electronics. These results confirm that solid-state based large-area detectors can achieve high-performance contamination monitoring while enabling modular, scalable, and easily deployable systems. The approach opens new perspectives for robotic applications and large-area inspection in nuclear environments.

Abstract

This work presents the experimental validation of SNIPER-GN, a portable gamma–neutron detection and identification system whose core innovation is a patented neutron identification algorithm (U.S. Patent No. 11835477). The algorithm enables robust and reliable neutron identification under challenging operational conditions, including shielding, moderation effects, and intense masking gamma radiation fields, representing a significant advancement over conventional detection approaches. The system integrates a high-efficiency gamma–neutron counter with a mid–high-resolution scintillation detector for rapid localization and characterization of radioactive threats, including Special Nuclear Material. The patented algorithm constitutes the enabling technology of the system and was the primary focus of the experimental verification campaign. Extensive performance testing was conducted in several international laboratories using calibrated sources under both static and dynamic measurement conditions. The system successfully identified neutron-emitting sources down to approximately 1 count per second with background levels below 0.1 counts per second, while gamma identification remained stable above 60 net counts per second over background. Tests employing polyethylene and lead shielding demonstrated identification success rates up to 100% depending on source geometry and activity, confirming robustness in realistic scenarios. Dynamic tests compliant with ANSI N42.53 confirmed reliable alarm performance for moving sources. Additional validation for field applications, such as rapid analysis of air-sampler filters and iodine cartridges for post-fallout monitoring, showed astonishing minimum detectable activities within measurement times shorter than five minutes.
Overall, these results validate the patented technology and demonstrate the suitability of SNIPER-GN for homeland security, emergency response, and radiological field deployment scenarios.

Abstract

Neutron imaging plays an irreplaceable role in the restoration of cultural relics and the study of lithium battery charging and discharging behaviors. The performance of neutron imaging detector directly determines the imaging quality. A cold neutron imaging detector based on a structured ¹⁰B₂O₃-ZnS:Cu scintillation screen was developed and tested at China Mianyang Research Reactor (CMRR) Cold Neutron Radiography Facility. The structured screen was made of glass microchannel plate (MCP), B₂O₃, ZnS:Cu and UV-curable binders. The refractive index contrast between the MCP (1.5) and the UV curable adhesive(1.62) leads to total internal reflection, thereby forming a light confinement structure. The spatial resolution was measured to be about 33μm with a Siemens Star Chart, and the cold neutron absorption efficiency was measured to be 24.4%

Abstract

SANDALS-II is an instrument being developed at the ISIS Neutron and Muon Source as a next generation diffractometer for the study of amorphous and liquid materials. This upgrade will deliver count rates up to 11 times higher than the existing SANDALS instrument and will significantly expand its scientific capabilities. This work focuses on developments of a large area scintillator detector that are essential for achieving enhanced performance. Scintillation detectors for SANDALS-II will cover a solid angle of 0.67Ω, 2.4 times greater detector coverage than the current instrument. These detectors are required to have high efficiency, particularly at higher neutron energies (short wavelengths), which is achieved using a unique design employing tiles made from alternating layers of glass and scintillator. Scintillator and glass options are explored to optimise detector and instrument performance. Different low Li and B content glasses were tested to determine the most suitable option for the instrument that provides good neutron transmission while maintaining affordability. The ratio of ZnS:Ag to ⁶LiF in the scintillator will be altered from 4:1 to 2:1 to enhance neutron absorption and therefore detection efficiency. At 30eV (0.05Å) this offers a 60% improvement in detection efficiency. The applicability of low-afterglow scintillator, ZnO:Zn/⁶LiF, for use in SANDALS-II is also explored. ZnO:Zn/⁶LiF is shown to offer a 2x increase in count rate capability over ZnS:Ag/⁶LiF, but a 6% lower neutron detection efficiency. As the reduction in efficiency using ZnO:Zn/⁶LiF is modest, this alternative scintillator offers a viable solution for the high count rate regions of SANDALS-II. With optimised glass and scintillator selection, the performance required for the upgraded instrument can be delivered. This will enable faster measurements, smaller sample studies, and time resolved experiments, making SANDALS-II a world leading disordered material neutron diffractometer.

Abstract

This study evaluated the performance of a LaBr₃(Ce) scintillation detector for elemental identification and mass-fraction estimation in a thermal-neutron-based prompt gamma-ray neutron activation analysis (PGNAA) system, in comparison with a conventional NaI(Tl) detector. PGNAA spectra were acquired at the Four-Circle Diffractometer (FCD) beamline of the HANARO research reactor at the Korea Atomic Energy Research Institute. The measured samples included pure Mn, Fe, Co, and Ni samples; binary Fe–Mn, Fe–Co, and Fe–Ni mixtures; and ternary NCM622 and NCM811 samples simulating spent lithium-ion battery cathode materials. Elemental identification was performed using Gaussian-derivative-convolution-based peak detection, SNIP baseline correction, and Currie-criterion-based significance testing. Mass fractions were estimated using region-of-interest-based weighted non-negative least squares (WNNLS) with attenuation-integration correction. LaBr₃(Ce) produced higher Z-scores, smaller energy-matching errors, and more stable characteristic peak combinations in mixed samples than NaI(Tl). For NCM622 and NCM811, LaBr₃(Ce) achieved mean absolute errors of 4.3 and 1.9 percentage points, respectively, while also providing higher R² values for spectral-shape reproduction. These results indicate that LaBr₃(Ce) offers improved peak separation and compositional estimation capability for nondestructive analysis of complex industrial samples, including spent lithium-ion battery cathode materials.

Abstract

This work presents a novel neutron detector concept aiming to provide detection efficiency and background levels equivalent to conventional ³He-based detectors, whilst boosting the maximum counting rate by several orders of magnitude. The detector retains the form-factor of a standard ³He tube and uses the same working gas mixture of ³He with admixture of CF₄ as the stopping gas. The key innovation is a fast optical readout of CF₄ prompt scintillation light using silicon photomultiplier (SiPM) sensors. The target performance is a detection efficiency equivalent to that of a standard ³He tube under identical gas filling conditions, gamma sensitivity below 10⁻⁶, and a maximum counting rate of 2 MHz/pixel, orders of magnitude improvement over existing detectors. The fast emission of CF₄ scintillation light is the enabling factor for achieving ultra-high counting rates. CF₄ exhibits a considerable photon yield, with one scintillation band in the UV and another in the visible region of the spectrum, with several thousands of photons emitted per neutron event. The efficient collection of light by the SiPMs constitutes a critical design parameter, addressed in this study through Monte Carlo optimization. The gas mixture composition is optimised to maximise the scintillation yield, and a gas purification system can be integrated into the detector to ensure long-term stability. The first experimental results of a detector prototype characterisation on a neutron beamline at the Institut Laue-Langevin (ILL) will be presented.

Abstract

The response of Timepix detectors with different sensors (Si, SiC, GaAs, CdTe) to thermal, slow, and fast neutrons is studied. Measurements at tunable mono-energetic neutron fields are compared and analyzed with nuclear-reaction models and numerical calculations. Specific radiation-type and sensor-material interaction neutron detection channels are derived for each detector-sensor configuration. These results serve to better and more completely process radiation data acquired with these detectors. Combining multi-detector measurements expands and enhances the resolving power of field decomposition.

Abstract

Neutron contamination in advanced radiotherapy and other mixed radiation environments has become an increasing concern due to the limitations of conventional neutron detectors under intense gamma background conditions. In this work, ultra-thin 3D silicon neutron detectors specifically designed for operation in mixed radiation fields were experimentally characterized under thermal and high-energy irradiation conditions. The detectors, consisting of 20 µm-thick silicon sensors paired with ⁶LiF neutron conversion layers, were tested at the FANT thermal neutron facility and at the CERN High-Energy Reference Field (CERF). Measurements included reproducibility, angular dependence, bias voltage operation, and discrimination threshold effects using a low-voltage FPGA-based readout system. The results showed stable operation under mixed radiation fields, moderate angular dependence, and strong dependence of the response on detector bias and discrimination threshold. The ultra-thin detector architecture enabled high gamma rejection capability, enabling reliable neutron detection under intense photon background conditions. These results demonstrate the suitability of 3D ultra-thin silicon detectors for compact real-time neutron monitoring in radiation protection and medical applications.

Abstract

In this document, a pileup deconvolution scheme not relying on any mathematics guessing is presented.  In high energy physics experiment, as the luminosity increases, pile-up issues on detectors such as calorimeters become non-negligible.  Deconvolution approaches developed for data taken from DAQ systems are usually rank-deficient or underdetermined, having less equations than unknowns, even with the ADC values from multiple beam crossings are collected. These deconvolution approaches need mathematic pre-assumptions such as Sparse Representation.  For online computation tasks such as for trigger primitive creation, signal availability is significantly different as in offline data analysis stage, and therefore, it is possible to use different (yet simpler) algorithms.  In this situation, number of ADC values of the calorimeter outputs is the same as the number of beam crossings (or 4 times number of beam crossings, depending on the ADC sampling rate), and therefore, the number of equations can be arranged to be the same as number of unknowns.  This way, a determined deconvolution scheme with a full-rank squared convolution (and deconvolution) matrix becomes possible.  The robustness of deconvolution over long time windows is also studied in this paper.

Abstract

The Neutron Radiation Detection Instrument-1A (NeRDI-1A) is a technology demonstration payload on the International Space Station (ISS) as part of the Department of Defense Space Test Program (STP) mission STP-H9. The primary objective of NeRDI-1A is to space-qualify two detector technologies: a dual-mode gamma-neutron  scintillator (Thallium-doped Cesium Lithium Yttrium Chloride, Tl₂LiYCl₆:Ce, or TLYC) and Domino microstructured semiconductor neutron detectors (MSNDs). The full NeRDI-1A suite includes a TLYC detector, three Domino MSNDs with varying moderation, and an EJ-270 plastic scintillator. This work focuses on the TLYC detector within NeRDI-1A. Here we present an analysis of the TLYC detector's response within distinct orbital regions--including the auroral regions and the South Atlantic Anomaly--during a period of minimal solar variability. We also present the evolution of the TLYC detector’s performance over its operational lifetime.

Abstract

The Assessing Lunar Ion-Generated Neutrons (ALIGN) payload is an advanced spectrometer developed to characterize the lunar radiation environment during the Chang'E-8 mission. Designed to operate within a stringent mass allocation of 4.8 kg and optimized for low electrical consumption, the payload features a highly integrated coincidence telescope statically oriented at -30 degrees from the horizontal. This presentation details the instrumental architecture and multi-particle detection methodologies of the payload. The detector array comprises Double-Sided Silicon Strip Detectors (DSSDs), an EJ-254 boron-loaded plastic scintillator, and a primary CLYC calorimeter, all enveloped by an active BC-408 plastic scintillator veto. The system employs the ΔE-E technique—correlating initial energy loss in the DSSD transmission trackers with total energy deposited in the scintillators—to separate upward-moving albedo protons (20–140 MeV) from background electrons. Concurrently, intrinsic Pulse Shape Discrimination (PSD) within the CLYC crystal is utilized to accurately separate epithermal neutrons from gamma-ray interactions. To achieve definitive particle identification within these strict size, weight, and power constraints, a dedicated onboard FPGA executes priority-based signal processing. Geant4 simulations evaluated the instrument's overall geometric acceptance as a function of incident kinetic energy. By combining these advanced onboard logic techniques with a streamlined composite radiation shield, structural design is expected to successfully isolate lunar albedo signals from localized lander interference while fully satisfying the mission's strict engineering constraints.

Abstract

The LHCb experiment recently underwent an extensive upgrade program to allow operation at higher luminosities. The VErtex LOcator (VELO) was redesigned to a hybrid pixel layout with 130nm CMOS technology. As a result of the upgrade, the VELO will be exposed to an extreme radiation environment – over an order of magnitude greater than before. At the end of its lifetime, the innermost regions are expected to be exposed to 8x10¹⁵MeV n_eq cm^-2 [1]. Previous analysis of bias voltage scans (presented at IEEE Yokohama, 2025) has shown the VELO is at risk of exceeding the maximum 1000V bias voltage limit to remain at operational performance. Here, we present a data-driven simulation of clusters in the VELO, to predict the performance at the end of its lifetime. Extensive validation of the simulation to testbeam data is performed, with angle and bias voltage scans showing very good agreement. Furthermore, charge sharing via diffusion in the simulation is authenticated with the charge collection vs intra-pixel hit position for size 2 clusters. Excellent agreement is seen. Comparison to Run III data is presented, showing good correspondence with the installed sensor. Extrapolations of the sensor performance at 8x10¹⁵MeV n_eq cm^-2 are performed. Preliminary results estimate the hit efficiency to be 89.41±0.14\% and the spatial resolution to be 19.98±0.37μm (a 19.8% increase from non-irradiated). These results highlight the need to make use of beneficial annealing for the rest of the sensor lifetime, to ensure performance remains as good as possible.

Abstract

To support the X-ray Imaging Telescope (XIT) onboard China's Solar Polar Orbiting Observatory, which aims to perform the first-ever imaging of the Sun polar regions and study high-energy processes in solar eruptions, we have developed a prototype readout electronics system for a CdZnTe hard X-ray detector. The system consists of a 2 mm thick CdZnTe detector with an effective area of 15 × 15 mm² segmented into four strip electrodes with switchable area ratios, a front-end electronics board employing the IDE4281 charge-sensitive preamplifier ASIC, and a FPGA-based data acquisition unit with integrated calibration and high-voltage supply. The IDE4281 chip provides 12 channels with programmable shaping times, self-triggering, and peak-hold outputs. In the readout prototype, the equivalent noise charge after common-mode noise subtraction ranges from 149 to 189 e⁻ across all channels. Using a ²⁴¹Am radioactive source, the detector demonstrates a full-width at half-maximum energy resolution of 4.92 keV at 59.5 keV, fulfilling the scientific requirement of better than 5 keV at that energy. These results validate the prototype design and establish the basis for the forthcoming X-ray imaging telescope instrument.

Abstract

Since its installation, the Timepix3 detector network has demonstrated utility within the ATLAS cavern by providing radiation field measurements and relative luminosity monitoring. This work evaluates the performance and internal stability of the network during the entire LHC Run 3. To ensure robust monitoring and assess systematic uncertainties, three distinct algorithms were evaluated and compared: cluster counting, thermal neutron counting, and heavy and straight track counting. 

Clean signal was achieved by identifying and masking noisy pixels and by modeling and subtracting background activation by a recursive growth and decay model. Internal consistency was verified through both short-term (intra-fill) and long-term (yearly) stability analyses across detectors situated at different distances from the interaction point. The presented results demonstrate that the distributed Timepix3 network provides a stable and robust relative luminosity measurement, complementing primary ATLAS luminometers during Run 3.

Abstract

The LHCb Upgrade II, planned for Run5, will operate at a substantially higher luminosity than previous runs, motivating dedicated FLUKA simulations of the resulting mixed-field radiation environment. This work addresses the impact of two major changes: the upgraded Electromagnetic Calorimeter, referred to as PicoCal, and the shielding of the Muon System. The increased fraction of high-density material in the PicoCal central modules modifies shower development and neutron production, while additional shielding is required to preserve Muon System performance. All comparisons are made with respect to the same Run4 reference configuration without additional Muon System shielding. Two Muon System shielding solutions are evaluated: i) replacement of the Hadronic Calorimeter by an iron-and-concrete shield with borated-polyethylene plates upstream and downstream, and ii) retention of the present Hadronic Calorimeter followed by a ductile cast iron, borated-polyethylene and steel wall. The simulations show increased local neutron production around the PicoCal central modules, while the existing polyethylene-and-boron shield upstream of the calorimeter limits the impact on upstream sub-detectors. For the Muon System, both shielding strategies reduce the charged-particle fluence relevant for detector performance, with different spatial and particle-composition trade-offs. These results support the optimization of the LHCb Upgrade II layout and illustrate the role of detector radiation studies for shielding design in mixed-field high-energy physics environments.

Abstract

FARCOS is a Femtoscope Array for Correlation and Spectroscopy designed to perform studies of two- and multi-particle and Intermediate Mass Fragments (IMFs) correlations in heavy-ion collisions at Fermi energies with stable and radioactive beams. The FARCOS system is a modular assembly of up to 20 telescopes to be used in combination with 4π detectors like the CHIMERA multidetector. Each telescope features two Silicon detection layers (32 + 32 channels Double Side Silicon Strip Detectors 300μm and 1500μm thick and a third calorimetric stage, made of four truncated pyramidal CsI(Tl) crystals arranged in a 2 x 2 matrix). Particular care has been posed in the design of the frontend electronics for the different stages. For the first time we performed a detailed qualification of the timing properties of the FARCOS frontend electronics and of the different detection stages. The achievable resolution was optimized thanks to an accurate analysis of the waveform shape. The remarkable time resolution measured at zero pf capacitive input load, with the FARCOS frontend electronics, below 100 ps rms at 1.1 MeV injected charge on the 100 MeV full-scale energy range, thanks to the excellent energy resolution below 15 keV FWHM paves the way to widen the range of experiments in which the FARCOS system will be applied. The measured time resolution improves to 25 ps rms at 5 MeV injected charge. The selectable gain factor allows the extension of the full-scale energy range up to a gain factor of 0.5mV/MeV over the full-scale range of 1.1V, with a small compromise on the time resolution, translating in a dynamic range of more than 15bits. The INL keeps below 0.5% over the full dynamic range. The timing performance of the Silicon layers of the FARCOS telescopes was then qualified with a near-infrared pulsed laser system and with alpha particles. This contribution presents the results of this qualification together with some highlights of the recent applications of the FARCOS telescopes.

Abstract

Silicon Photomultipliers (SiPMs) are solid-state photodetectors increasingly employed in astrophysics and space missions due to their high photon detection efficiency (PDE), compactness, insensitivity to magnetic fields, and operation at relatively low bias voltages. However, exposure to radiation, through both Ionizing Energy Loss (IEL) and Non-Ionizing Energy Loss (NIEL) effects, can degrade their key performance parameters. 
To mitigate radiation-induced damage, annealing strategies have been explored. This work presents the results of two annealing tests performed on FBK SiPMs, after x-ray irradiation and proton irradiation, highlighting their respective advantages and limitations: (1) thermal annealing, involves storing the devices in a climatic chamber without active heating, while periodically performing electrical measurements to track recovery over time. (2) In-situ current annealing progressively increases the SiPM bias current to induce self-heating via resistive effects. Electrical measurements have been done after each step, and a thermocamera have been used to monitor device temperature.

Abstract

This work presents the measurement of the axial neutron emission profile of the Polaris field-reversed configuration (FRC) fusion machine using CR-39 solid-state nuclear track detectors (SSNTDs). Passive detectors were deployed at multiple axial positions along the Polaris compression section, accumulating neutron-induced proton recoil tracks over several pulses of moderate yield, sufficiently low to resolve tracks. Chemical etching in 6M NaOH at 85°C for 2.5 hours revealed latent damage tracks, which were analyzed using a custom Python-based image processing pipeline incorporating a trained Random Forest classifier achieving 94.1%  ± 2.2% cross-validation accuracy in neutron track identification. The resulting axial track density profile is symmetric and centrally peaked, in agreement with MCNP-simulated neutron dose distributions for a thermonuclear DT fusion source confined within the FRC plasmoid. The absence of asymmetric structure - characteristic of beam-target dominated reactions - provides compelling experimental evidence of spatially confined thermonuclear fusion on Polaris. This represents the first deployment of CR-39 diagnostics for reactivity profiling on a pulsed FRC machine and establishes passive track detectors as an effective tool for fusion burn characterization.

Abstract

The ATLAS Liquid Argon (LAr) calorimeter readout electronics are being upgraded for operation at the High-Luminosity Large Hadron Collider (HL-LHC), where up to 200 proton-proton interactions per bunch crossing will significantly increase detector occupancy, trigger rates, and radiation levels. While the calorimeter detectors themselves remain unchanged, the complete readout electronics chain is being replaced. The upgraded on-detector electronics readout system continuously amplifies, shapes, digitizes, and transmits calorimeter signals from approximately 182000 channels at the 40 MHz bunch-crossing rate. The new Front-End Boards (FEB2) integrate custom radiation-hard ASICs including the ALFE2 preamplifier/shaper, the COLUTA analog-to-digital converter, and lpGBT optical communication devices. Prototype boards and ASICs have been developed and extensively tested in laboratory and integration environments. This contribution presents the architecture, development status, and recent performance results of the Phase-II front-end electronics for the ATLAS LAr calorimeter.

Abstract

The first generation of planar high-speed 2D charge-integrating imaging detectors is widely used in scattering experiments at free-electron laser facilities. Although diffraction is inherently defined on a spherical geometry, planar detectors are typically employed to record diffraction patterns. Current MHz frame rate detector systems offer spatial resolutions limited by pixel pitches in the range of 200 μm to 500 μm . Next-generation devel- opments aim to achieve sub-100 𝜇m resolution, introducing new challenges. At these smaller pixel sizes, second-order effects such as charge sharing, parallax distortion, and geometric projection errors become increasingly significant, whereas they are largely negligible at coarser resolutions.

In this work, we analyze the fundamental limits of distortion- free imaging imposed by planar detector geometries specifically for charge-integrating detectors, and explicitly excluding single-photon-counting approaches. We show that, for X-ray photon energies between 6 keV and 25 keV, distortion-free measurements are restricted to scattering angles of 2Θ = 0° and ≲ 13° for a 50 μm pixel pitch and 2Θ = 0◦ and ≈ 30° for 100 μm pixel pitch. Beyond these angular ranges, geometric distortions dominate the effective spatial resolution, such that a detector with 50 μm pixel pitch provides little resolution advantage over a 100 μm system. Our results identify fundamental scaling limits of planar charge-integrating detector concepts for scattering applications and establish general design constraints relevant for future high- resolution large-area detector developments at X-ray light sources.

Abstract

Back illuminated, HPGe, microstrip sensors are operated in the cryogenic detector system called XH. XH enables the collection of energy dispersive X-ray absorption spectra to investigate near-instantaneous structural variations in experimental samples. XH detects the transmitted radiation of a bright, single flash of X-rays that is delayed with respect to a laser pulse, which also hits the sample and creates an excited state of matter. Each strip in the HPGe sensor detects the intensity of a spectral component (∆E ∼1.5 eV) of the transmitted, dispersed polychromatic beam, which has a typical bandwidth of ∼1 keV. The HPGe sensor can be exposed to beam intensities as high as 1 × 10¹³ ph/mm²/s from polychromatic beams with a central energy in the range ∼5 keV to ∼25keV. The operation in this radiation harsh condition benefits from a systematic irradiation campaign to investigate ageing effects created by Total Ionising Dose (TID). The TID effects on the HPGe sensor are mitigated by design by adopting a back illumination approach in order to shield the strip contacts, and by the selection of a proprietary, commercial lithographic process. XH was irradiated at the X-ray Irradiation Facility at Daresbury Laboratory (UK). The 1024 strip sensor was repeatedly exposed to X-rays in finite time steps. Total leakage and strip currents were monitored regularly with X-rays on and off. The X-ray peak energy was limited to 20 keV, which was fully absorbed in the 1.5 mm thick HPGe sensor. This protected the strip contacts from radiation. Beam collimation within the active area of the HPGe sensor showed a steady performance, which was immune from TID. Evidence of deterioration in surface leakage current was measured when the beam was hitting the sensor edge and its supporting PCB.

Abstract

White beam Laue diffraction is well-known as a means to orient single-crystal samples, but its use as a structure-determining tool is less exploited. One of the difficulties with the method is its incomplete measurement of the diffraction vector. Only its direction can be determined, but not its length, since conventional instrumentation does not provide the diffracted beam energy. In this paper we propose to use a high-resolution energy resolving detector to provide that information, together with NSLS-II’s HEX beamline, which has a spectral range up to 200keV. We have developed a series of microstrip germanium detectors, and present here some measurements using one of them together with a horizontal aperture to emulate a 192 x 1 pixel detector with 250 micrometer pixels. By scanning this device in one dimension we could simulate a 2-dimensional detector with good energy resolution,. We have further designed and built a custom detector for this purpose, which has 192x2 pixels, each 250 x 250 micrometers, and so does not need the aperture, and which requires half the number of measurements. Data collected with these detectors will be presented.

Abstract

A fully digital readout architecture is described for high-throughput X-ray spectroscopy with silicon drift detectors (SDDs). The core challenge is the large-amplitude voltage ramp produced by transistor reset preamplifiers, which conventionally prevents direct digitization at spectroscopic resolution. In this work, sampling the preamplifier output with a 20-bit SAR ADC achieves a dynamic range exceeding 10⁶, eliminating any analog conditioning stage upstream of the digitizer.

Throughput is sustained by combining four converters in a time-interleaved scheme, yielding an aggregate sampling rate of 160 Msps. A Zynq SoC provides the processing backbone, performing inter-channel timing alignment, gain equalization, and concurrent execution of the signal processing pipelines.

Two energy estimation algorithms run in parallel on the digitised data: a ramp-corrected trapezoidal filter and a model-based least-squares fitter that uses linear regression over the full inter-event sample record, remaining accurate even when pulse separation shrinks to sub-microsecond intervals. Experimental characterisation with ARDESIA multi-pixel SDD arrays and KETEK single-pixel detectors yields an energy resolution of 120.7 eV FWHM at low rate, 150 eV at 1 Mcps, and 180 eV at 2 Mcps, with live time exceeding 90% throughout.

Abstract

Fluorescence-mode X-ray Absorption Spectroscopy (XAS) at high photon energies require detectors with both high stopping power and excellent energy resolution to measure weak element-specific fluorescence signals. Conventional silicon-based detectors suffer from poor absorption efficiency above ~20 keV, while most high-𝒁 semiconductor materials such as cadmium telluride (CdTe), cadmium zinc telluride (CZT), and gallium arsenide (GaAs) have largely been developed for imaging applications with comparatively limited spectroscopic performance. In contrast, Germanium combines excellent stopping power with superior intrinsic energy resolution owing to its low Fano factor and favorable charge transport properties, making it particularly attractive for spectroscopy in the 15-100 keV range. Here we report the development, fabrication, and characterization of Gepta-EX, a compact multi-channel high-purity germanium (HPGe) detector system designed for fluorescence-mode XAS. Gepta-EX incorporates a monolithic seven-channel HPGe pixel array integrated with low-noise CUBE charge-sensitive preamplifiers inside a thermally isolated compact cryostat operating near 90K. The detector achieves energy resolutions of 218 eV at  5.9 keV using an ⁵⁵Fe source and 373 eV at 59.5  keV using an ²⁴¹Am source, with uniform response across all channels. By avoiding the escape-peak interference commonly associated with silicon detectors while maintaining stable high-resolution performance at elevated X-ray energies, Gepta-EX provides a powerful platform for high-energy X-ray spectroscopy and fluorescence-mode XAS applications.

Abstract

At nuclear facilities such as decommissioning sites and reprocessing plants, continuous monitoring of radioactive liquid waste is essential to prevent the spread of contamination. While gamma-emitting radionuclides can be detected relatively easily using gamma-ray detectors, radionuclides emitting only alpha or beta radiation are difficult to measure remotely because these particles are strongly absorbed in water. Conventional liquid scintillation counting requires sample collection and pretreatment, making real-time monitoring difficult. In this study, a real-time visualization method for beta-emitting nuclides in water was developed using a calcium fluoride (CaF₂) scintillator and an Electron Multiplying CCD (EM-CCD) camera. The scintillator was designed with a water-permeable structure to allow contaminated water to flow through it, enabling direct detection of beta radiation in water. Radiation source measurements were performed in a dark room using a 90Sr–90Y beta source. Scintillation light generated in the CaF₂ scintillator was captured by the EM-CCD camera. The results demonstrated that the position of the beta source could be successfully visualized with acquisition times as short as 1 s, and the detected intensity increased with measurement time. Measurements conducted while moving the source confirmed that changes in source position could also be visualized, indicating the capability to monitor moving contamination in water. In addition, water infiltration tests confirmed that water could successfully pass through the scintillator structure. These results demonstrate that the proposed system is promising for real-time monitoring and visualization of beta-emitting radionuclides in radioactive liquid waste and aquatic environments.

Abstract

This paper presents a methodology for measuring the trajectories of high-energy electrons through materials utilizing a transmission electron microscope (TEM). This novel technique is implemented to diagnose and quantify the mechanism for enhanced radiation attenuation via nanostructures. Rather than an imaging technique, the TEM is applied to radiation shielding measurements as an adaptable electron source with functionality for various beam energies, amperage, alignments, and exposure times alongside an array of precise electron detectors. By utilizing energy filtering and direct electron detection capabilities, scattering profiles are measured with minimal background contributions. This TEM method is leveraged to describe particle transport in nanostructures. Palladium nanoparticles at various monodisperse sizes are synthesized via heat injection and deposited into an epoxy matrix. The palladium epoxy nanocomposites are sectioned with an ultramicrotome to thicknesses of 70 to 1000 nm containing nanoparticle sizes of 5 to 50 nm at 20% of the total weight fraction. Scattering profile measurements at various beam energies are conducted with a variety of sample thicknesses, nanoparticle sizes, and composite fractions. Scattering profile measurements also include bulk, homogeneous palladium samples in contrast to nanomaterials composed of identical atomic percentages. Simulation of the experimental scattering profiles demonstrate an understanding for the phenomena behind nanostructured attenuation enhancement, bolstering advanced nuclear technology via a foundational description for particle transport in nanostructures.

Abstract

New generation of vertex detectors must have excellent thermo-mechanical stability, and all-silicon ladders are considered as promising solutions for future detectors. However, power consumption asymmetries between the active area and periphery can induce large temperature gradients, and in some cases a combination of air and liquid cooling will be required to achieve homogeneous thermal profiles. Microchannel cooling offers a competitive alternative to traditional cooling systems. The short thermal path between the heat load and sink gives an increased thermal evacuation performance, combined with the reduced material needed for the integrated microchannels and the reduced Coefficient of Thermal Expansion mismatch between the cooling and sensor materials to obtain an optimized cooling for radiation detectors. The development of the technology needed for the integration of the sensor with the microchannels in a single piece is an ambitious challenge, as the conditions for the microchannel formation must be compatible with the sensor, requiring the optimization of the fabrication to ensure the compatibility with CMOS processes. As a demonstrator, all-silicon ladder prototypes with integrated cooling microchannels are presented. Power consumptions for the active area and periphery of the ladders are emulated using integrated aluminum heaters. The bulk in the active area is thinned from 600 to 50 microns, and the design of the microchannels cooling circuit, integrated in the periphery of the ladder, is optimized using finite-element simulation packages. Variations of prototype ladders with different sizes, microchannel dimensions, and integrated using different wafer bonding techniques are fabricated and evaluated in detail in this work.

Abstract

We present performance calculations for a proof-of-concept, spin-selective neutron detector in which the magnetic field of a microfabricated planar superconducting coil controls the quantum mechanical absorption resonance in a multilayer structure. For first tests with cold neutrons, depolarization calculations favor a parallel over a meander geometry, the latter of which should nevertheless offer enhanced detector sensitivity in the ultracold neutron regime. Simulations predict that spin contrasts can be efficiently demonstrated using polarized cold neutron reflectometry at existing facilities.

Abstract

Metallic magnetic calorimeters can provide excellent energy resolution for decay-energy spectrometry of alpha- and beta-emitting radionuclides, but temperature-dependent gain drift can broaden peaks during long acquisitions. This work compares two post-acquisition correction models for data acquired with a gold-erbium paramagnetic sensor in the MAGNETO-ν neutrino experiment. The pre-trigger baseline is used as an event-by-event proxy for detector temperature. The lower-order model includes linear baseline dependence, whereas the higher-order model adds baseline-energy cross terms and second-order baseline dependence to capture nonlinear drift. Model parameters are estimated with a likelihood-based fit to known plutonium and americium reference peaks, and each event is calibrated by solving the fitted response model for deposited energy. Performance is evaluated from isotope-resolved peak widths in the corrected alpha-decay region of the spectrum. Across the reference groups considered here, the higher-order model produces consistently narrower fitted peaks, with the clearest improvements in the higher-statistical-weight isotope groups. These results indicate that higher-order baseline terms can improve spectral resolution when baseline-correlated drift is present.

Abstract

We present a low-energy electron beam position detector based on a segmented capacitive sensing architecture designed for precision monitoring of electron beams in the 500 eV to 2000 eV energy range. The detector employs a 3 mm × 3 mm capacitive sensor structure optimized for non-invasive beam position measurements with micrometer scale spatial resolution. Low-energy electron beams present unique diagnostic challenges due to their sensitivity to electromagnetic perturbations, beam spreading, and reduced signal levels. Conventional beam diagnostics often suffer from limited sensitivity or introduce disturbances that alter the beam trajectory.

The proposed detector addresses these limitations through a pixelated capacitive pickup array combined with low-noise front-end electronics and differential signal processing techniques. Variations in induced charge distribution along the sensing electrodes are used to reconstruct the transverse beam position with high precision while maintaining minimal interaction with the electron beam. The array enables accurate position tracking over extended beam paths and improves sensitivity to small beam displacements.

Analytical modeling and preliminary experimental characterization demonstrate that the sensor can resolve beam motion at the micrometer level across the targeted energy range. The detector response was evaluated as a function of beam energy, beam current, and electrode geometry, showing stable operation and high linearity under low-energy conditions. Noise mitigation strategies, including shielding and optimized capacitance matching, were implemented to preserve signal integrity and reduce environmental interference.

This detector architecture is well suited for applications requiring compact, high-resolution diagnostics of low-energy electron beams, including micro cathode ray magnetic probes, electron optics experiments, and vacuum electronic devices.

Abstract

Reliable long-term monitoring in nuclear decommissioning facilities is hampered by the absence of permanent power infrastructure and by the radiation fields that degrade conventional electronics. We present a scintillator-photovoltaic converter that captures ambient gamma radiation and transforms it into electrical energy sufficient for periodic wireless data transmission. The converter pairs a 750 cm³ CsI(Tl) crystal with a compact GaInP photovoltaic array mounted on a single end face, following a "large absorber, minimal semiconductor" strategy.) On-site evaluation inside a contaminated-material storage building at the Fukushima Daiichi Nuclear Power Plant (~35 mSv/h) produced 273 nW per module. Two parallel modules, combined with ultra-low-power energy management and a sub-GHz radio, achieved seven autonomous transmission bursts totaling 32 packets over a 30 m link during a three-hour observation window—without any external power supply or dedicated radioactive source.

Abstract

The Metal-Hydride and Gas (MEGA) detector is a recently proposed approach for high-selectivity tritium (T₂) monitoring that combines hydrogen (H₂) storage metals with gas proportional counting. Previous work demonstrated the feasibility of T₂ absorption in palladium (Pd), formation of metal-tritide, and subsequent T₂ release into a proportional counter for measurement. In this work, we investigate pathways toward improving the overall detection efficiency of the MEGA detector through optimization of T₂ absorption, retention, desorption, and counting characteristics. A simplified efficiency framework is introduced to identify the dominant efficiency-limiting processes. The pressure–composition behavior of the Pd–H system is examined to evaluate approaches for enhancing isotope uptake at low T₂ partial pressures through reduced operating temperature and increased local pressure. Preliminary Monte Carlo simulations of tritium β-particle transport in Pd indicate strong self-absorption effects, with the simulated maximum penetration depth of 18.6-keV β-particles in Pd limited to a few hundred nanometers. These results support the need for optimized Pd geometry and efficient thermal desorption approaches to maximize gas-phase T₂ release and proportional counting efficiency. The present work establishes a quantitative framework for future optimization of MEGA detector sensitivity for environmental and nuclear safeguards applications.

Abstract

Polarization observations in extreme ultraviolet (EUV) band have not yet been achieved because it remains technically difficult to develop wide-field EUV detectors with sufficient sensitivity. However, EUV polarization measurements are expected to provide important insights into the emission mechanisms of gamma-ray bursts (GRBs) and ultraviolet flashes associated with supernova explosions. In this study, we propose a novel EUV polarimeter based on Lobster-Eye Optics (LEO), which combines a wide field of view with polarization sensitivity. Optical simulations show that EUV reflectivity on metal surfaces strongly depends on polarization. For example, at 40 eV and an incident angle of 75° on a gold-coated mirror, the reflectivity is about 40% for p-polarization and 60% for s-polarization, respectively. This difference produces an asymmetric cross-shaped image on the focal plane detector depending on the polarization direction. To verify this principle, we developed a test chamber and conducted a synchrotron radiation experiment using a gold-coated mirror and a microchannel plate (MCP) detector. By rotating the mirror around the beam axis, we measured the modulation of the MCP count rate and evaluated the detector modulation factor. The experimental results are presented and discussed in this paper.

Abstract

The KNU Advanced Positronium Annihilation Experiment (KAPAE) Phase II searches for invisible positronium decays using a compact 4π BGO scintillation calorimeter. Stable operation, efficient photon collection, and low-background performance are essential for this rare-event measurement. We report the cryogenic and underground performance of a chemically diffused BGO scintillator array developed for KAPAE Phase II.

The BGO crystal surfaces were chemically diffused to enhance photon collection and coupled to SiPM-based readout for low-temperature operation. At approximately −37 °C, reduced SiPM thermal noise and improved light collection yielded an energy resolution of 9% at 662 keV. The scintillation decay time was approximately 960 ns. This performance improves upon earlier BGO-based positronium calorimeters, including the ETH Zurich experiment, which reported an energy resolution of approximately 15% [1].

The detector was installed and operated at Yemilab underground. A continuous 120-day run demonstrated stable detector response, trigger operation, and long-term data-taking capability under cryogenic conditions. Calibration data were used to tune and validate a Geant4 detector simulation, which was applied to evaluate energy containment, gamma-ray leakage, and the expected sensitivity for invisible para-positronium decay searches. Based on measured performance and validated simulations, KAPAE Phase II achieves the expected sensitivity of 10−8 for invisible para-positronium branching-ratio measurements.

This study demonstrates that chemically diffused BGO scintillator arrays can provide stable, high-resolution calorimetric performance under cryogenic underground conditions.

Abstract

Six commonly used models for neutron induced light yield in organic scintillators were fit and analyzed using measurements of a 2.54 cm diameter by 2.54 cm long right circular cylinder of EJ301D and a (5.08 cm)3 custom made cube of deuterated trans-stilbene-d12 (d-stilbene). Measured neutron energies spanned 300 keV to 2200 keV for the EJ301D scintillator and 200 keV to 2500 keV in d-stilbene. The model fits and experimental analysis methods were verified by examining the agreement between the predicted light output spectrum and the measured light output spectrum for several simulated quasi-monoenergetic neutron sources. Birks model inaccurately predicted highly quenched light yield for sub-MeV neutrons in both scintillators. The addition of a second order quenching term to the Birks model by Chou significantly improved the models accuracy, however its behavior is non-physical if extrapolated beyond the measured neutron energies. The model derived by Voltz accurately fit the measurement of both scintillators for all measured neutron energies. Further investigation into the ability to extrapolate based on these semi-empiric models would be valuable given the non-physical behavior of well-fitting models.