Online program Overview Session: RTSD-04

Abstract

Accurate dosimetry of high-energy (≥ 15 MeV) clinical electron beams is challenged by bremsstrahlung-induced photon contamination (X-ray tail), which leads to unintended dose deposition in deep tissues. In this study, high-resistivity CdZnTe (CZT) detectors were investigated for sensitive characterization of this contamination. Two planar CZT detectors (5-mm and 8-mm thick) exhibited high bulk resistivity (~2.0 × 10¹⁰ Ω‧cm) and competitive energy resolution (3.9–4.1% at 662 keV). Measurements were performed using clinical LINAC electron beams (6–18 MeV) at an SSD of 100 cm. To circumvent pulse pile-up under intense photon flux, the detector response was evaluated via time-integrated current in a mixed radiation field. The integrated charge increased markedly at 15 and 18 MeV, indicating enhanced bremsstrahlung production. The 8-mm detector showed higher sensitivity to penetrating photon components than the 5-mm detector, highlighting the importance of detector thickness. These results demonstrate that CZT detectors enable high-sensitivity characterization of photon contamination and are promising for real-time dosimetry and radiotherapy quality assurance.

Abstract

CsPbBr₃ single crystals (CPB SCs) were grown by a modified inverse temperature crystallization method and processed to fabricate metal-insulator-semiconductor (MIS) sensor structures. A 5 nm Al₂O₃ layer was deposited by atomic layer deposition to passivate the crystal surface, suppress surface leakage, and provide environmental protection. Pt, Ti, Au, and Al contacts were subsequently deposited to form asymmetric barrier structures with different electrode combinations. X-ray diffraction measurements confirmed high crystalline quality of the as-grown CPB crystals. The rocking curve of the (004) plane exhibited a full width at half maximum of 0.19°. I-V measurements demonstrated pronounced rectification in all investigated MIS structures. The Pt/Al₂O₃/CPB/Al₂O₃/Al sample exhibited the lowest reverse current (~ 25 nA at 1000 V) and the strongest blocking behavior. These results confirm that Al₂O₃ passivation combined with optimized asymmetric contacts is an effective approach for engineering low-leakage, polarization-resistant perovskite X/γ-ray sensors.

Abstract

This study presents a systematic investigation of additive engineering strategies for the solution-based growth of all-inorganic Cs₂AgBiBr₆ double perovskite single crystals and their evaluation as radiation detectors. Various chemical additives including toluene, ethyl acetate, thiourea, sodium acetate (NaOAc), and potassium acetate (KOAc) were systematically introduced into the precursor solution to influence nucleation dynamics, crystal size, and structural quality. The combination of NaOAc and KOAc was identified as the most effective additive strategy, consistently producing large (up to 2.2 cm), well-faceted, phase-pure single crystals with high reproducibility. The optimized crystals were characterized using X-ray diffraction(XRD), high-resolution XRD(HR-XRD) with rocking curve analysis, scanning electron microscopy(SEM), energy-dispersive X-ray spectroscopy(EDX), laser-induced breakdown spectroscopy (LIBS), and Raman spectroscopy. HR-XRD rocking curves revealed exceptionally narrow full-width at half-maximum (FWHM) values as low as 0.0144°, confirming outstanding crystalline perfection. Electrical characterization, including current–voltage (I–V) profiling, space-charge-limited current (SCLC) analysis, signal-to-noise ratio (SNR) evaluation, was performed to assess the detector potential of the optimized crystals. While the crystals exhibited high bulk resistivity (∼10⁹–10¹⁰ Ω·cm), excellent phase purity, and radiation sensitivity, detector-level investigations revealed trap-limited charge transport and self-trapped exciton formation as intrinsic limitations. This work establishes dual-acetate additive engineering as a robust strategy for growing high-quality Cs₂AgBiBr₆ single crystals and provides critical insights into the gap between crystal quality and detector-scale performance.

Abstract

This work builds upon our characterization of lead-free double perovskite Cs₂AgBiBr₆ single crystals for radiation detection applications, where high-quality, centimeter-scale crystals (up to ~2.2 cm) were synthesized via a low-cost standard solution-cooling growth method. Structural analysis confirmed a phase-pure cubic Fm3̅m structure with a dominant {111} orientation and narrow FWHM values, indicating high crystalline quality. Morphological and compositional studies revealed smooth surfaces, uniform morphology, and near-stoichiometric composition, while electrical measurements showed low leakage current, ohmic behavior, and high resistivity (10⁹–10¹⁰ Ω·cm)[4]. Despite these favorable properties, device performance remained limited due to defect-related charge transport.To address this, a constant-temperature, feed-controlled growth approach is introduced, where supersaturation is regulated via continuous precursor feeding under stable thermal conditions. This steady-state method suppresses uncontrolled nucleation and promotes the growth of larger, higher-quality crystals. The approach is expected to reduce defect densities and improve material performance, providing a pathway toward high-performance, environmentally friendly radiation detectors.

Abstract

In this work we determined the peculiarities of high-temperature phase equilibria in the CdTe–MnTe system using a differential thermal analysis method in a thermocycling mode. Temperature dependences of the volume fractions of the liquid and solid phases in Cd_1-xMn_xTe (x=0÷0.5) alloys, as well as the minimum and maximum melting temperatures of the solid phase were obtained. These dependences allowed us to establish critical temperatures of equilibria between the solid phase and the melt and to construct the solidus and liquidus lines of the CdTe–MnTe phase diagram. It was shown that with an increase in the molar fraction of MnTe in alloys ranging from 0.1 to 0.5, the decreases in the temperatures for both the solidus and the liquidus did not exceed 20 °C. Plus, the temperature interval between them remains practically constant at 15-18°C. Also, we determined the temperature intervals of the pre-melting region and found that with an increase in the MnTe fraction, the region of pre-melting states of the alloys expands. The results described in this work provide more reliable data on the equilibrium between phases, since the method reduces the influence of kinetic limitations of classical thermography.

Abstract

To fully exploit the high-precision of proton therapy, advanced beam monitoring and quality assurance devices capable of reliable real-time dose verification are essential. This study presents the first performance evaluation of a Gallium Nitride (GaN) pin-diode array at a clinical proton therapy facility using Pencil Beam Scanning. The detector array is comprised of 11x11 diodes with a 500 μm pitch coupled to a low-noise multi-channel readout system, and the response of all 121 pixels was evaluated individually. The experimental characterization was carried out at a cyclotron-based proton therapy center using clinical proton energies of 100 MeV and 226.7 MeV. We investigated the detector response for different single proton spots ranging from 0.02 to 100 monitor units (MU), corresponding to 10⁶ - 10¹⁰ protons. The results demonstrate excellent linearity between diode response and machine output. Regarding the time response characteristics, a bi-exponential decay model expresses the observed signal fall-off after irradiation. The afterglow contribution was less than 5% of the integrated signal per measured spot. These initial results of a pixelated GaN pin-diode array under clinical pencil beam scanning conditions identify the proposed detector as a promising candidate for both real-time beam monitoring and high-precision quality assurance in clinical proton therapy environments.

Abstract

All-inorganic perovskite CsPbBr₃ single crystals are promising candidates for room-temperature radiation detection; however, the conventional vertical Bridgman (VB) growth rates (<1 mm/h) severely constrain efficient growth and wide applications. In this study, we report a novel strategy for preparing high-performance single crystals via zone refining, which integrates the traditionally separate purification and crystal growth processes into a unified in-situ purification and growth process. This strategy accelerates the growth rate to 25 mm/h — approximately 25-fold faster than conventional vertical Bridgman (VB) growth methods — while preserving an X-ray rocking-curve FWHM of just 0.02°. The resulting crystals exhibit exceptional carrier transport properties, with a hole mobility-lifetime product (μₕτₕ) of 1.2 × 10^-3 cm² V^-1 and a dark current of only 26 nA at 300 V for a 6-mm-diameter circular device. Asymmetric planar detectors fabricated from these crystals demonstrate outstanding room-temperature energy resolutions (ERs) of 7.3%, 5.8%, 5.0%, and 3.3% for ²⁴¹Am (59.5 keV), ⁵⁷Co (122 keV), ²²Na (511 keV), and ¹³⁷Cs (662 keV) γ-rays, respectively. This work breaks the crystal growth rate bottleneck and provides a practical pathway toward the scalable in-situ growth of high-performance perovskite CsPbBr₃ radiation detectors.

Abstract

Photon-counting detectors (PCDs) enable high spatial resolution, zero electronic noise, and multi-material decomposition. Chromium-compensated gallium arsenide (GaAs:Cr) is an outstanding candidate material for these applications. To address single-crystal inhomogeneities and ASIC-induced variations, we implemented a single-pixel energy calibration scheme based on K-edge absorption for high-resistivity GaAs:Cr detectors [1]. S-curve scans were performed for 0.6 mm, 1.0 mm, and 1.4 mm sensor thicknesses using 0.5 mm metallic filters (Ta, W, Pb). Calculating the differential intensity ratio between filtered and air spectra allowed accurate K-edge energy extraction, yielding highly linear fits (R ² > 0.99) between energy (keV) and Digital-to-Analog Converter (DAC) thresholds. Crucially, incorporating electron mobility-lifetime (µτ_e) products elucidates the physical mechanisms driving thickness-dependent gain variations and spectral shifts. Subsequent evaluations confirmed the calibration accuracy. This calibration and physical analysis validate the spectral imaging capabilities of thick GaAs:Cr detectors, advancing their clinical applications.

Abstract

This work investigates the polarization and depolarization processes in CdZnTe and CdTe radiation detectors with rectifying contacts. The electrical properties of the contacts were characterized by I–V measurements, while the evolution of space charge was studied using the Laser-induced Transient Current Technique (L-TCT) combined with Monte Carlo simulations of charge transport to determine electric field profiles. In both CdZnTe and CdTe detectors, the gradual formation of negative space charge within the detector volume after applying bias was observed due to the hole blocking anode. Depolarization methods based on optical irradiation were subsequently investigated. In the CdTe detector, continuous above-bandgap irradiation of the anode by a 565 nm diode resulted in a stable, nearly flat electric field profile, demonstrating effective suppression of polarization for at least 1000 s. Similar results were observed for above-bandgap anode irradiation of CdZnTe as well. In both CdZnTe and CdTe detectors, continuous side irradiation by a below-bandgap 904 nm diode decreased the space charge and improved stability as well. However, compared with the above-bandgap anode irradiation, polarization was only partially suppressed for the CdTe detector, and the internal electric field still evolved over time, albeit more slowly than in the dark. For the CdZnTe detector, the side irradiation completely stopped the polarization. The results confirm that controlled optical irradiation can significantly improve the stability of CdTe/CdZnTe detectors operated under bias.

Abstract

Conventional 4H-SiC Schottky barrier detectors (SBDs) with thick active layers (>50 µm) are often limited by high leakage currents and premature breakdown under high bias voltages. This work introduces an innovative approach using an ultrathin yttria-stabilized zirconia (YSZ) high- κ dielectric interlayer to form a metal-oxide-semiconductor (MOS) structure. While YSZ is a known ionic conductor at elevated temperatures, its ionic mobility is effectively frozen at room temperature, allowing it to function as a robust electrical insulator that suppresses carrier injection. Ni/YSZ/n-4H-SiC MOS structures, fabricated via pulsed laser deposition demonstrated stable operation up to 1 kV. The high-voltage stability enabled a depletion width of approximately 70 µm at 1000 V, establishing favorable conditions for efficient charge collection in thick epitaxial layers. Spectrometric evaluation with a triple-alpha source (²³⁸Pu–²³⁹Pu–²⁴⁴Cm) confirmed enhanced performance, with stable energy resolution maintained across a broad bias range (100–1000 V) and optimal results observed between 200 V and 600 V. This architecture significantly improves high-voltage resilience while maintaining competitive spectroscopic qualities, paving the way for next-generation deep-depletion radiation detectors in nuclear and space applications.

Abstract

Abstract— TlBr Polycrystalline Based Semiconductors (PBS) are investigated as promising ceramic materials for low-cost, high-energy x-ray detection. Ceramic samples were prepared through hot pressing followed by the fabrication of planar detectors for characterization and x-ray response measurements. Current-voltage (I-V) characteristics were evaluated for all devices, which exhibited high resistivity (≥10⁹ Ω cm) and stable performance under applied bias. The samples demonstrated excellent sensitivity to 70 kVp x-rays, as well as outstanding response to 6 MeV and 9 MeV x-ray pulses from the Linatron M9A linear accelerator (linac) at the University of Tennessee, Knoxville (UTK). This work provides a general overview of the ceramic preparation process and detector characterization and summarizes low and high-energy x-ray responses measured at RMD and UTK.

Abstract

This work investigates the influence of Schottky contact size on the charge collection efficiency (CCE) and lateral expansion of the active region in 350 µm thick semi-insulating GaAs detectors using alpha particle spectrometry. Measurements were performed for detectors with Schottky contact diameters ranging from 0.10 mm to 2.00 mm under reverse bias voltages up to 500 V. The results show that detectors with smaller contacts achieve higher CCE values and simultaneously higher counting rates, indicating spreading of the electric collection field beyond the geometrical contact area. The lateral extension of the active region was evaluated from the effective active diameter derived from alpha peak count analysis. The obtained results demonstrate an approximately linear increase of the lateral extension with increasing reverse bias voltage, reaching average values of about 80 µm at 200 V and 120 µm at 300 V. These findings are particularly relevant for pixelated detector systems with small pixel pitch, such as Timepix3-based detectors employing thick GaAs sensors.

Abstract

Understanding charge cloud formation and transport in CdTe detectors is essential for precision spectroscopy, imaging, and Compton polarimetry in the sub-MeV gamma-ray band. We report a detailed experimental study of electron diffusion and electron-dispersion mechanisms in a 2.0 mm-thick CdTe-Timepix3 detector (256×256 pixels, 55 µm pitch), operated at a bias voltage of −300 V and temperature of ≈28 °C. Monochromatic 200 keV and 300 keV gamma-ray beams, collimated to 10 µm×10 µm, were provided by the ESRF ID15A beamline and directed along the detector’s Y-axis. A depth-resolved scan was performed in 200 µm steps along the Z-axis, probing the evolution of the transverse charge cloud inside the CdTe volume. For each Z-position, thousands of photoelectric events were selected, and the 1D transverse distribution of the total collected charge was reconstructed along the X-axis. Gaussian fits to these distributions yielded depth-dependent width values. Fitting the squared width values with a function that models all contributions to the transverse charge distribution at the pixelated plane enabled the extraction of key charge-transport parameters: the initial radius of the primary photoelectron cloud and the electron transverse diffusion coefficient. Independent measurements of the electron drift velocity using throughgoing muon tracks were used to constrain this parameter. The measured evolution of sigma values is well described by a model that combines thermal diffusion, the finite initial radius of the primary photoelectron cloud, and Coulomb-driven charge-cloud expansion. This work provides the first depth-resolved experimental quantification of these combined effects in CdTe-Timepix3 detectors, improving charge-transport inputs for simulations of Compton imaging and polarimetry in the 0.1–1 MeV energy band.

Abstract

Methylammonium lead tribromide (MAPbBr₃) single crystals are intensively studied emerging materials for use in X-ray radiography and gamma-ray spectrometry [1] and their quality such as shape control, dislocation and point defect densities are improving but some of their properties remain elusive. In particular, transient X-ray-induced photocurrent can be measured in MAPbBr3 single crystals detectors after the bias is switched off. This response then slowly decays over time. In this study, we investigate the internal electrical field that builds up when a bias is applied2⁴ by measuring the residual sensitivity of millimeter-thick Cr/MAPbBr₃/Cr-based detectors after switching off the bias from 10V to 0V. By fitting a numerical model, we compute the value of this internal field and infer the ionic charge density close to the electrodes. This model is also used to approximate the charge carrier mobility lifetime product.

Abstract

High‑performance γ‑ray spectroscopy with solution‑grown metal halide perovskites has been persistently limited by uncontrolled defect formation during crystal growth. Here, we identify ambient mechanical vibration as a critical yet overlooked factor that destabilizes the concentration boundary layer and perturbs lattice ordering in hybrid perovskite single crystals. The resulting crystals exhibit an increased carrier mobility-lifetime product of 8.36×10^-2 cm^-2 V^-1, which contributes to a more efficient charge collection process. Detectors achieved an energy resolution of 9.5% at 59.5 keV and 5.2% at 122 keV. These results establish vibration control as a key principle for high‑quality single‑crystal growth and scalable perovskite radiation detectors.

Abstract

The study of the electrical properties of perovskite CsPbBr₃ crystals, grown by the Bridgman method in our laboratory, showed that the stability of the dark current of the In/CsPbBr₃/In structure critically depends on the polarity of the applied voltage. The CsPbBr₃ crystal was cut perpendicular to the growth plane. Indium was deposited on both surfaces after treatment in an argon plasma. The contact area was 4x4 mm², and the thickness of the plates was 1-2 mm. The study was carried out at a temperature of T≈300 K and a voltage of 100 V. The surface of the sample, which was closer to the beginning of the ingot growth and crystallized first, will be designated as “cold”. When on the “cold” side was the “minus”, the current remained stable over time for 8 hours of continuous operation, while with reverse polarity, a degradation of the blocking properties wais observed with an increase in the dark current by 2–3 orders of magnitude. A similar phenomenon (non-equivalence of opposite crystal surfaces) was observed in CdTe crystals, oriented in the (111) plane. Therefore, it can be assumed that when cutting a plate perpendicular to the growth axis, with there is a very high probability that one side will be enriched with a layer of Cs and Br atoms, and the opposite side, which is more chemically active, will be enriched with Pb and Br atoms. This may be the reason for the non-equivalence of their behavior in the electrical measurements.

Abstract

Perovskite single crystals demonstrate great potential in high energy radiation detection applications due to the unique semiconductor properties, while still suffers from uncontrollable growth progress and labile crystal quality. In this paper, we here develop a series of interdisciplinary growth methodologies for fabricating high-quality perovskite single crystals based on the solvent design and additive engineering, achieving high-performance and reliable robustness toward impressive X-ray detection performance.

Abstract

Methylammonium lead tribromide (MAPbBr₃, MPB) single crystals are promising candidates for direct-conversion X-ray detection, but their practical use is hindered by ion migration that causes photocurrent drift and baseline instability under continuous X-ray exposure. In this work, a two-dimensional (2D) organic–inorganic hybrid layer is formed on MPB single crystals through a simple solution-dipping process using 4-fluorophenethylammonium iodide (4F-PEAI), and the impact on detector stability is systematically investigated. The 2D-layer formation, controlled through precursor concentration and dipping time, is confirmed by the emergence of a new low-angle XRD peak assigned to a (4F-PEA)₂PbX₄ 2D layered phase, and by a blue shift of the band-edge photoluminescence consistent with surface modification. Surface 2D-layer passivation reduces the dark-current hysteresis, suppresses the photocurrent drift under continuous X-ray exposure, and lowers the in-pulse relative standard deviation, while maintaining linear dose response across the measured dose-rate range. These results demonstrate that surface 2D-layer formation is a practical strategy for suppressing ion migration and improving the operational stability of perovskite single-crystal X-ray detectors.

Abstract

Real-time proton beam monitoring is crucial for pencil beam scanning in proton therapy, but conventional solid-state detectors suffer from radiation-induced damage and structural rigidity, limiting their in-vivo applications. Building upon the lattice-anchoring dynamic repair mechanism in hybrid perovskites, we developed a highly stable, flexible perovskite thin-film transistor array for real-time proton beam profiling. The detector adopts a p-i-n architecture on a flexible substrate, protected by robust Parylene encapsulation to ensure long-term operational stability. Under 160 megaelectron-volt high-energy proton irradiation, the arrays demonstrated exceptional radiation hardness, maintaining robust operational stability even after accumulating a high fluence of 1.23*10¹⁰ p cm^-2. Spatial resolution remained uncompromised post-irradiation, with the spatial frequency at a modulation transfer function of 0.2 showing negligible degradation from about 3.03 line pairs per millimeter.. When deployed in a clinical setup, the TFT array achieved real-time proton beam imaging at a frame rate of 15 frames per second. At a 16-centimeter solid water depth, the measured beam full-width at half-maximum (FWHM) was 8.47 millimeters, perfectly matching the 8.27 millimeters theoretical value predicted by Monte Carlo simulations using FLUKA and Allpix, and it has achieved real-time tracking and prediction of the center of the proton beam. These results validate the immense potential of flexible perovskite TFT arrays for the real-time tracking of the Bragg peak and conformal dose monitoring, offering a transformative dosimetric tool for advanced precision radiotherapy and expanding the practical applications of room-temperature semiconductor detectors.

Abstract

Formamidinium lead bromide (FAPbBr3) single crystal perovskites have recently emerged as a key candidate for commercially viable radiation detectors for their high per formance and solution processability at low temperatures. An under-explored consequence of crystal growth from solution is the ability to crystallise around pre-placed structures, enabling detectors with an embedded electrode architecture. This work presents the growth of FAPbBr3 single crystals around a 2×2 array of gold wire electrodes, showing the first demonstration of this device architecture in a FAPbBr3 single crystal radiation detector as a proof of concept. By collecting charge carriers laterally, the charge carrier transit time and trap probability is reduced compared to the planar configuration. Initial charac terisation demonstrated a linear I–V response and sensitivity to X-rays. Future work will scale to a larger array with an increased area targeting γ-ray spectroscopic performance with fast signal pulse rise times.

Abstract

Thallium bromide (TlBr) is a highly promising material for next-generation gamma-ray detectors due to its wide band gap and high stopping power. To realize large-volume and high-performance detectors, it is essential to establish an efficient evaluation method for detector performance. This study aims to investigate the correlation between crystalline structural quality and carrier dynamics in TlBr single crystals. We developed a scanning time-of-flight (TOF) measurement system to evaluate the two-dimensional distribution of electron mobility and trapping characteristics. A 3-mm-thick TlBr wafer was characterized using both EBSD mapping and laser-induced TOF measurements. The results revealed a clear correlation between structural grain boundaries and localized degradations in electron mobility. Furthermore, spatial mapping of the residual charge ratio identified specific regions characterized by high trap densities. Subsequent evaluations of detectors fabricated from these regions showed significantly poor long-term stability. This study provides essential feedback for improving crystal growth processes and establishes a methodology for the quality assurance of compound semiconductor radiation detectors.

Abstract

This study describes the fabrication of the aramid nanofiber (ANF) to produce flexible detectors based on nanomaterials, with short preparation and low surface leakage current. Flexible detectors offer the advantage of achieving improved detection efficiency and efficient use of space with curved surfaces. In previous studies, ANF was used as a porous nanostructured polymer for the fabrication of nanomaterial-based semiconductor detectors. However, it had the issues of a long preparation time and brittleness after drying. To address these issues, we used DMSO containing an aqueous KOH solution in the preparation of the ANF solution. with This reduced the preparation time from one week to two days. Furthermore, we compared the results of ANF coagulation in different soloutions—ethanol (EtOH), acetic acid (AcOH), and a mixed AcOH-EtOH (volume ratio was 7:3) solution—to alleviating the pore collapse of the nanostructure, with  the conventional H₂O-based method. The synthesis solution containing PbS quantum dots (about dia. 5 nm) and Te nanorods (130 nm in length), as confirmed by TEM and XRD, was loaded into ANF samples using various coagulation solutions. We confirmed that uniform loading in dpeth occurred in the order of H₂O, EtOH, AcOH, and the AcOH-EtOH mixture. Furthermore, according to the I–V curves, the surface leakage current also tended to decrease in the same order of uniform loading, with a maximum 10,000-fold difference observed between the ANF samples prepared with H₂O (about 100 μA at 10 V) and the AcOH-EtOH mixed solution (about 2 nA at 10 V). Therefore, it was shown that ANF coagulated with the AcOH-EtOH mixed solution remained flexible even after drying, and the uniform embedding of nanomaterials according to depth reduced the surface leakage current.

Abstract

Perovskite materials are promising for low-cost direct X-ray detection, but their integration with thin‑film transistor backplanes is limited by porosity, cracking, and poor imaging uniformity. Here, we report a combined strategy of uniform particle dispersion and solution-assisted hot pressing for fabricating uniform MAPbI₃ thick films on TFT substrates, where uniform particle dispersion stabilizes slurry dispersion and solution-assisted hot pressing promotes film densification, grain growth, and pore filling. The average particle size of the fabricated devices decreased from 32.07 μm to 15.14 μm, and the standard deviation reduced from 7.3 μm to 4.9 μm. At an electric field strength of 100 V·mm^-1, the sensitivity reached 3514 μC·Gy_air^-1 cm^-2. A prototype MAPbI₃-based X-ray flat-panel detector further achieves a spatial resolution of 3.75 lp mm^-1 at an modulation transfer function of 0.2, together with significantly improved imaging uniformity. This work provides a practical route toward scalable, thin‑film transistor compatible perovskite X-ray flat‑panel detectors.

Abstract

CdTe is a well-established semiconductor for high-resolution X-ray, gamma-ray detector.  However, its widespread application is limited by high dark current and long-term performance instability. The metal–semiconductor interface plays a critical role in governing these issues, as it strongly influenced by contact metal indiffusion and interfacial reactions. In this work, indium tin oxide (ITO) is investigated as a potential contact material for CdTe detectors to mitigate these limitations. ITO electrodes were deposited on one face of a high-resistivity p-like CdTe crystal by RF sputtering, while a gold electrode was formed on the opposite face by thermal evaporation. The fabricated device showed clear rectifying behavior, with reverse dark currents below 1 nA at 100 V bias for a 3 x 3 x 0.5 mm³ device, representing a significant reduction compared to similar devices with Au contacts on both sides. Improved gamma-ray detection performance was also observed. Furthermore, the effects of CdTe substrate temperature during ITO deposition and post-deposition annealing on junction formation and detector performance, including dark current, energy resolution, and long-term stability, were systematically investigated.

Abstract

This study investigates the characterization of conductive layers on CdTe semiconductor crystals fabricated via laser ablation. The layers were prepared at both ambient and elevated temperatures, with various ablation parameters evaluated to optimize their quality. Our analysis revealed the formation of conductive layers on the semiconductor surface, exhibiting high hole mobility. These layers were systematically characterized using a suite of electrical and optical techniques, including Hall effect measurements, photocurrent mapping, the Laser-induced Transient Current Technique (L-TCT), and stability tests focusing on time-dependent behavior.

Abstract

Chromium-compensated gallium arsenide (GaAs:Cr) is
a highly promising material for spectral X-ray photon counting
detectors (PCDs), offering high electron mobility (∼4000 cm2/V · s),
low K-shell fluorescence (9.2 keV and 10.5 keV), and ultra-high
bulk resistivity (∼109 Ω · cm). A critical factor governing the high-
flux stability of such PCDs is the metal–semiconductor contact.
While conventional detectors often rely on Schottky contacts—which
trap space charge and cause polarization under sustained irradi-
ation—the ultra-high resistivity of GaAs:Cr enables barrier-free
Ohmic contacts. This approach relies entirely on bulk resistance to
limit dark current, eliminating interface charge accumulation and
preserving a uniform electric field. In this work, we characterize
the wide-bias (−1000 V to 1000 V) dark current–voltage (I–V )
behavior of pixelated GaAs:Cr detectors. The devices exhibit highly
linear, symmetric responses with distinct Ohmic and space-charge-
limited current (SCLC) regimes, confirming high-quality Ohmic
contact formation. Furthermore, the extracted bulk resistivity and
demonstrated pixel-level uniformity validate the contact integrity
required for stable high-flux photon counting applications.

Abstract

A Geant4-based Monte Carlo framework was developed to predict the proton-irradiation stability of FAPbBr3 perovskite. Beyond calculating non-ionizing energy loss (NIEL), the framework records layer-resolved energy deposition across a 2 mm crystal divided into 50 slices. For 70, 100, 150, and 200 MeV protons, the NIEL values are 2.04, 2.01, 1.38, and 1.55*10^-3 MeV cm^2 g^-1, confirming non-ionizing displacement-damage channels. The deposited-energy fractions in the incident-side 0.20 mm are 11.83%, 10.69%, 19.57%, and 14.90%, respectively, following the same order as the measured electrical response: 150 MeV > 200 MeV > 70 MeV > 100 MeV. This agreement indicates that near-surface energy deposition can effectively predict mobility-lifetime product variation and supports the framework as a screening tool for perovskite radiation stability.

Abstract

The effect of the Low Earth Orbit (LEO) environment on future high energy instrument sensors is essential to optimize their performance, by minimizing orbital background noise. In the framework of the CNES-ESA European Material Ageing program (EMA), the ageing effects on critical components of a Laue lens telescope in the LEO space environment is investigated. Since more than a decade, the group has been developing CZT drift strip sensors to meet the stringent requirements of a Laue lens focal plane detector for high-energy photons. As part of the EMA proposal, two simplified samples of drift strip CZT detectors were exposed to outer space on the CNES Bartolomeo external platform of ISS for more than 12 months. The objective was to test the robustness and reliability of the surface passivation and metallization processes in the LEO space environment. The samples were taken down to earth on February 27. In this work, we report on the results of the simulations we carried out accounting for possible damages to the detector properties due to agents as oxygen and high energy ions. Also, preliminary measurements on the detectors come back from space will be shown.

Abstract

We developed a compact CdTe-based gamma-ray spectroscopy system with a dual-side readout architecture for high-resolution gamma imaging. The system is designed for pixelated Schottky CdTe detector arrays. Cathode pixels at the same relative position across multiple detector chips are multiplexed, while the anode of each chip is read out independently. This configuration reduces the number of readout channels and system complexity while preserving energy information from the anode side and position information from the cathode side. 
The system integrates custom JFET-input charge-sensitive preamplifiers and a waveform-sampling data acquisition system based on AFE5801 chips, enabling measurement of pulse height, risetime, and waveform information.
To evaluate the feasibility of the proposed method, spectral measurements were performed using a 2 × 1 CdTe detector array operated at +650 V with sealed ²²Na and ¹³⁷Cs sources. A 1 μs coincidence time window was applied to match the anode and cathode signals. Calibrated energy spectra were obtained from both sides. The anode signal showed slightly better energy resolution and was used as the primary signal for energy measurement, while the cathode signal can provide position information. After applying risetime selection, energy resolutions of 3.61% at 511 keV and 3.20% at 662 keV were achieved. These results demonstrate the feasibility of the proposed dual-side merged readout architecture.
A 5 × 5 CdTe detector array has also been assembled, and further verification experiments are in progress.

Abstract

Three-dimensional position-sensitive CdZnTe (3D-CZT) gamma-ray detector systems provide compact gamma-ray spectroscopy with 3D position sensing. Achieving high performance requires event-by-event energy and position reconstruction to compensate for charge loss, gain variation, and response nonlinearity. Such reconstruction calculations are typically performed on a powerful CPU. To reduce power consumption and minimize system footprint, we propose implementing the reconstruction in an application-specific integrated circuit (ASIC). In this work, we focus on single-pixel event reconstruction, integrating depth and nonlinearity corrections along with a sub-pixel position calculation. This ASIC is implemented in TSMC 65 nm CMOS technology and is designed to interface with a DAQ-DSP back-end ASIC and four H3DD-UM front-end ASICs. Preliminary results demonstrate the feasibility of an event reconstruction ASIC with an area of 2.43 mm^2. This development represents a key step towards the development of more compact 3D-CZT systems.

Abstract

XRPIX is a silicon-on-insulator (SOI) CMOS pixel detector developed for next-generation X-ray astronomical observations. An X-ray camera, SUIM (Soipix for observing Upper atmosphere as Iss experiment Mission), is our first space mission employing XRPIX and is planned for launch to the International Space Station (ISS) in 2026. Thermal analysis of the SUIM instrument indicates that the camera housing may reach temperatures as high as +60°C in orbit, making it essential to verify XRPIX performance under elevated temperature conditions. The primary concern under such conditions is the degradation of spectroscopic performance due to increased dark current. Spectroscopic measurements using ²⁴¹Am were conducted over a temperature range from −30°C to  +70°C, with a periodic reset of 100 μs applied to suppress dark current accumulation during integration. Only minimal degradation was observed, confirming that the periodic reset scheme effectively maintains good spectroscopic performance even at +70°C. The gain decreases at a rate of −0.132 μV e⁻¹ °C⁻¹, which can be corrected through gain calibration. These results demonstrate that XRPIX maintains sufficient spectroscopic performance at least up to +70°C, indicating that its performance is also adequate for use in the SUIM instrument.

Abstract

 TlBr is a promising compound semiconductor material for gamma-ray detectors[1] because of its wide bandgap, high-Z constituents, and high density, which enable room-temperature operation and high detection efficiency. However, because the hole mobility in TlBr is significantly lower than the electron mobility, single-polarity charge sensing techniques are required to achieve high energy resolution. In this study, a zero-grid-bias operation in a coplanar-grid electrode structure for TlBr gamma-ray detectors was investigated. Unlike conventional coplanar-grid detectors, the proposed structure operates without a bias voltage between adjacent grid electrodes, which may reduce leakage current.

A conventional comb-shaped electrode structure was evaluated experimentally and through simulations using a Cs-137 source. The simulation results reproduced the experimental pulse-height spectra well. In addition, pulse-height correction using the cathode-to-anode pulse-height ratio improved the energy resolution

Abstract

This study investigates the spectral performance of a unipolar charge-sensing amorphous selenium detector using Monte Carlo simulation and compares its response with that of a conventional amorphous selenium detector. The unipolar charge-sensing architecture mitigates the limitations associated with poor electron mobility and deep electron trapping in amorphous selenium by promoting hole-dominated signal formation. Both unipolar charge-sensing and conventional detector structures were simulated for a 20 μm pixel pitch under an average electric field of 8 V/μm. The simulated full-width at half maximum energy resolution was 8.02 keV for the unipolar charge-sensing detector, compared with 21.32 keV for the conventional detector. These results demonstrate that the unipolar charge-sensing architecture substantially improves the spectral performance of amorphous selenium detectors by suppressing depth-dependent signal variations and reducing sensitivity to electron trapping.

Abstract

We report on the characterization of the Timepix4 hybrid pixel detector with a Si sensor under exposure to heavily ionizing particles. The device was configured sequentially in three gain modes: logarithmic (log), high, and low. Previously developed methodology of pixel's energy calibration beyond calibration with X- and $\gamma$-rays was adopted and applied to Timepix4 data. It uses the protons with energies in the range from 500 -- 2000 keV. Saturation of the pixel's energy response was observed at 1057, 1244, and 2030~keV in the low, high, and log gain modes, respectively. In a laboratory experiment, the detector was irradiated with $\alpha$-particles from a $^{239}$Pu$^{241}$Am$^{244}$Cm source, even higher energy $\alpha$-particles were obtained from naturally occurring radon progeny decays recorded over several weeks of continuous operation. Both measurements yielded an energy resolution of about 2.5\%. The capability of Timepix4 for relativistic ion beam decomposition was investigated within fragment beams at the Super Proton Synchrotron (SPS) at CERN. A spectrum stripping technique based on iterative Landau curve fitting was applied to the measured energy-loss spectra, enabling ion charge identification up to $Z$ = 10 at a track inclination of $\theta$ = 60 deg. with respect to the sensor normal.

Abstract

CdZnTe (CZT) detectors require stable calibration parameters for optimal spectroscopic performance. However, certain pixels exhibit random gain variation during or between measurements, complicating calibration and degrading detector performance. Digital ASICs allow for readout of event waveforms in a CZT detector exhibiting gain instability. Th-228 and Cs-137 measurements were utilized to characterize the gain instabilities, revealing a decrease in the anode amplitudes proportional to the incident energy. Analysis of the neighbor pixel-anode waveforms reveal abnormally high tail amplitudes in the four side neighbors which exceed expectations based on the Shockley-Ramo theorem. The corner neighbors experience noticeable smaller increases in tail amplitudes. Sub-pixel position analysis demonstrates that the tail amplitude increase is not proportional to the neighbor signal amplitude. These findings suggest that the observed gain instability could be caused by issues with capacitance in the system rather than anomalous charge movement. Ongoing work aims to verify the cause of the shift and if the anode deficit can be entirely explained by the differences in the neighbor waveform signals.

Abstract

In the present work, we study the MiniPIX Timepix hybrid pixel detectors with silicon sensors of two different thicknesses and their response to high energy protons. Thermally stabilized detectors measured protons of five energies, 400, 700, 1000, 1400 and 1800 keV in vacuum. During the measurements, a bias scan was performed to investigate the efficiency of collection of charge carriers generated by the protons. With increasing reverse bias, the signal from the detector grew until it saturated. The 300 um thick detector required an electric field generated by a reverse bias of 30 to 40 volts for effective charge collection. A detector with a thicker 500 um sensor required approximately 70 V bias for complete charge collection when registering 400 keV protons and a gradually higher bias of up to 100 V in the case of 1800 keV protons. The much larger bias for a complete collection of the charge generated by protons in a thick detector is in correlation with the theory, where the semiconductor depleted region thickness increases proportionally to the square root of applied reverse bias.

Abstract

The Timepix family of pixel detectors provides high spatial resolution together with per-pixel spectroscopic response, allowing event-by-event analysis of radiation interactions through deposited energy and cluster morphology. In fast neutron fields, the detector response is indirect and is produced mainly by secondary particles generated in the sensor and surrounding materials. This work focuses on fast-neutron-induced tracks measured with a Timepix3 detector equipped with a 300 µm silicon sensor at the Van de Graaff laboratory of IEAP CTU in Prague, using neutrons produced trough D–T reaction, in the range of approximately 15–17.5 MeV. The data were processed using TraXEngine [1] and further evaluated through selected spectral and morphological cluster parameters. The selected parameters were analyzed in two- and three-dimensional parameter spaces, where high-density regions corresponding to characteristic event groups were identified. Based on these regions, five parameter-based classes were defined: high-energy ion-like events, low-energy ion-like events, a transition region toward electron-like events, photon-like events, and electron-like events. The developed parameter-space analysis shows that combinations of spectral and morphological parameters can be used to separate components of a fast-neutron-induced mixed radiation field and can support future automated particle-recognition methods.

Abstract

    Direct-conversion X-ray flat-panel detectors are attractive for high-resolution and low-dose imaging compared to indirect conversion. However, conventional passive pixel sensors (PPS) without in-pixel gain are commonly used in direct-conversion detectors, causing low signal-to-noise ratio under low-dose exposure. In this work, a compact two-transistor active pixel sensor (2T-APS) with in-pixel current amplification based on amorphous indium-gallium-zinc-oxide (a-IGZO) dual-gate TFTs is proposed for low-dose direct-conversion X-ray imaging. The proposed pixel employs two subthreshold-biased dual-gate TFTs to realize logarithmic–exponential signal processing, enabling electrical tuning of the operating point and sensitivity.
     A 512×256 2T-APS array was fabricated using an a-IGZO dual-gate TFT process with a pixel pitch of 75 μm×75 μm. Pixel-level characterization further demonstrated amplification of pA-level input currents, achieving a representative current gain over 10^3. The fabricated array was integrated with a perovskite layer to realize X-ray imaging. At the tube voltage of 50 kV, images acquired with tube currents from 0.5 mA to 2.0 mA clearly visualized the object outline. These results validate the feasibility of the proposed a-IGZO dual-gate 2T-APS pixel for low-dose, high-sensitivity direct-conversion X-ray flat-panel detectors.

Abstract

Simultaneous multi-nuclide imaging is a key challenge in nuclear medicine, promising benefits such as improved diagnostic accuracy, reduced examination frequency, and elimination of physiological variations between scans. However, conventional imaging systems such as positron emission tomography (PET) and single-photon emission computed tomography (SPECT) are based on fundamentally different principles, making their integration into a single system technically difficult. To address this, we have proposed and developed a Compton-PET camera that enables simultaneous visualization of PET and SPECT nuclides using PET and Compton imaging, respectively. In our previous implementation, the system employed GAGG scintillators as absorbers, but their limited energy resolution (11–15% FWHM above 100 keV) degraded the Compton imaging performance. Thallium bromide (TlBr) semiconductor detectors have been shown to offer both high energy resolution, such as 2% FWHM at 662 keV, and high time resolution, including 511 ps via Cherenkov photon detection, making TlBr a promising sensor material for improving Compton-PET performance. Based on these advantages, we developed a new multi-pixel TlBr semiconductor detector by bonding a TlBr crystal to a self-triggering, energy-resolving pixel application-specific integrated circuit (ASIC). We plan to present the results of I-V characteristic measurements and evaluate detector performance through energy spectrum measurements and time resolution analysis using radionuclide sources.

Abstract

The INFN SPECTRE (Single Photon Emission Computed Tomography Readout Electronics) project aims to design, assemble, and characterize a complete detection module based on CZT sensors for medical applications. As the demand for SPECT systems with higher spatial resolution, energy resolution, and sensitivity continues to grow, there is an increasing need to develop high-performance CZT detectors with small pixels and corresponding high-speed readout electronics [1,2].
For readout, SPECTRE uses electronics developed for a high-capacitance gas detector (CGEM@BESIII, IHEP, Beijing [3]): the TIGER (Torino Integrated GEM Electronics for Readout) ASIC for on-detector electronics and the off-detector FPGA-based GEMROC module for power and data processing [4]. TIGER is a mixed-signal chip with 64 channels on an area of 5 × 5 mm². Each channel has a complete readout chain with amplification, signal conditioning, and discrimination, as well as a data payload containing the time and charge information for each event. TIGER is versatile for the readout of radiation sensors up to 50 fC and for high rates up to 70 kHz. The TIGER/GEMROC readout circuit is compact, modular, scalable, and highly customizable.  In typical CZT applications today, spectroscopic systems operate at rates of about 1 kcps/mm² because extracting energy information requires more processing time [5]. However, our system can acquire each sensor pixel at a rate of up to 70 kHz, or approximately 16 kcps/mm².  
In this presentation, we will cover all phases of the project to date, including CZT sensor and electronics simulation with the Allpix Squared package [6], based on Geant4, current-voltage characterization of the CZT sensors, system design and construction, assembly of four modules onto an adapter board, and initial system characterization using X-rays.

Abstract

Photon-counting computed tomography requires readout ASICs with high count-rate capability and reliable spectral performance under clinical X-ray flux. In this work, a peak-finding ASIC for GaAs:Cr photon-counting detectors is presented. The prototype ASIC contains a 9 × 12 pixel array with a pixel pitch of 341 μm × 368 μm. Each pixel integrates a charge-sensitive preamplifier, shaper, peak-finding circuit, five energy bins, threshold DACs, and counters. The peak-finding circuit captures the amplitude of signal peaks, improving counting capability under high-flux conditions. Electrical measurements demonstrate reliable counting performance beyond 30 Mcps/pixel. After coupling with a GaAs:Cr sensor, X-ray measurements showed clear tungsten characteristic peaks, confirming the energy-discrimination capability of the detector module. These results demonstrate the potential of the proposed ASIC for high-flux GaAs:Cr-based PCCT detectors.

Abstract

Photon-counting detectors (PCDs) with high-Z sensors are widely used for spectroscopic X-ray imaging in medical, security, and industrial applications. Hybrid PCDs can be coupled to different semiconductor sensors (Si, CdTe, GaAs), tailoring the assembly to the application. CdTe is the sensor of choice for energies above 50 keV, given its higher absorption efficiency compared to Si. The Timepix3 is a pixelated hybrid PCD that records each individual photon interaction in the sensor, providing the time-over-threshold (energy) and time-of-arrival of each event at fluxes up to 40 MhitsHits/cm²/s. The spectral resolution of CdTe-Timepix3 assemblies degrades with sensor thickness, and the relative weight of front-end electronic noise and sub-threshold charge-sharing losses has not yet been quantified at 55 µm pitch. An empirical equivalent-noise-charge (ENC) model exists for Si-Timepix3 but, to our knowledge, has not been reported for CdTe at this pitch. This work addresses this gap by measuring the front-end ENC of three CdTe-Timepix3 assemblies (1 mm, 3 mm, and 5 mm sensor thickness) from internal test-pulse scans, fitting an empirical ENC model with the detector capacitance estimated from per-pixel geometry, and decomposing the photopeak FWHM at the ²⁴¹Am (59.5 keV) line into Fano, front-end, and sub-threshold contributions. A Si-Timepix3 of the same ASIC family is included as a cross-check for the front-end term. The result is a quantitative basis for predicting CdTe-Timepix3 spectral performance as a function of sensor thickness on this ASIC.

Abstract

In semiconductor radiation detectors such as silicon (Si) and silicon carbide (SiC) the sensor bias determines the radiation sensitive volume. Varying the sensor applied bias can thus significantly change and even distort the detector response, sensitivity and spectral-tracking properties. These effects can be studied and described when the sensor is bump-bonded to the Timepix3 detector which provides particle counting response, high-granularity imaging detection and spectral-tracking ability to single particles. These properties make possible to analyze and resolve particle-type events which is useful for mixed-field decomposition. In this work we examine and experimentally investigate the response of the SiC Timepix3 detector to well-defined fast neutrons to quantify and analyze the changes induced by varying the sensor applied bias voltage. The main effects are further decreasing the detection efficiency, which is overall low (≤ 1%), and only partly reducing the spectral-tracking resolving power to particle-type discrimination.

Abstract

This study presents an experimental and computational framework for gamma-ray imaging using a position-sensitive Cadmium Zinc Telluride (CZT) detector combined with deep learning–based reconstruction techniques. A GDS-100 CZT detector with an 11 × 11 pixelated anode matrix was irradiated using a tightly collimated 662 keV Cs-137 gamma-ray beam scanned across the detector surface with 1 mm step size. Detector waveforms were used to train a compact ResNet-based convolutional neural network capable of reconstructing higher-resolution spatial information by mapping 11 × 11 detector inputs to 22 × 22 outputs. Kernel density estimation (KDE) analysis of reconstructed events shows that the predicted interaction positions remain strongly localized around the true coordinates, demonstrating accurate spatial learning. Average spatial spreads of σx = 1.375 mm and σy = 1.398 mm, together with positional deviations below 1 mm, indicate stable and nearly symmetric localization performance. Although edge and corner regions exhibit slightly broader distributions due to detector-related physical effects, the reconstructed interaction densities remain well confined around the true positions. Overall, the results demonstrate that deep learning–based CZT signal analysis can significantly improve spatial reconstruction performance in semiconductor gamma-ray imaging systems.

Abstract

Reliable radioisotope identification with handheld instruments remains challenging due to limited detector resolution, shielding, low counting statistics, and variable background conditions. We present a convolutional neural network algorithm for automated multi‑isotope identification from gamma‑ray spectra acquired with low‑resolution detectors. The algorithm is trained using an optimized synthetic spectral library of 48 radionuclides and evaluated using measured data acquired under controlled shielding, signal‑to‑noise ratio, and count‑rate conditions. Results show near‑perfect identification accuracy for moderate to high signal‑to‑noise ratio regimes and for low to high statistics. At low signal‑to‑noise ratio the algorithm is still able to recognize some radionuclides with decent accuracy. These results demonstrate that the method provides a practical and scalable solution for real‑time isotope identification in nuclear security and radiation detection systems.

Abstract

In the context of in-situ radiological characterization of sites (natural or artificial radioactivity) and for the dismantling of nuclear installations, high-resolution gamma-ray spectrometry is a powerful tool to both identify and quantify gamma-emitters. This technique makes it possible to better define the state of radiologically polluted soils without having to collect samples. To perform mapping of large areas (a few hundred square meters) and to detect radiological hotspots, we first use a 2’’x2’’ NaI(Tl) scintillator probe (Mirion® SG2R) at 20 cm from the ground in pedestrian mode, called C.R.E.M.P (Caractérisation Radiologique En Mode Piéton, radiological characterization in pedestrian mode). Then, for each particular points, high-resolution gamma spectrometry is performed with an HPGe Ortec® DETECTIVE-X. To obtain a reliable transfer function, the absolute efficiency of the detectors must be known. We therefore used our radiological characterization platform located at CEA/DAM Ile de France. Thanks to these measurements, a geometric model of the HPGe detector was developed and optimized using the MCNP Monte Carlo simulation code taking into account real position and shape of the germanium crystal including mechanical supports and dead layers. To validate the model, experiments on radiologically marked standard soils were performed. Extrapolated data from MCNP simulations make it possible to obtain the transfer functions for each detected radionuclide. The whole method, measurements and simulations will be described.

Abstract

This paper contains the findings associated with the collection of gamma spectroscopy of a variety of fuel types being utilized in the deployment of the next generation of nuclear reactors. Gamma spectroscopy of TRISO fuel types at HALEU levels of enrichment were collected and analyzed prior to installment in their respective reactor designs. Preliminary measurements of these fuels establish a strong baseline that serve as important experimental data points for simulation validation and monitoring fuel characteristics that are underexplored in next generation reactor designs such as burnup rates, and fuel stability (physical, chemical, performance).

 In addition, the established baseline also serves as an important benchmark for quality assurance and nuclear nonproliferation by providing reference spectrum for archiving purposes in a nuclear fuel landscape that is rapidly expanding. This ensures enrichment reporting is accurate in the event of fuel mixing/reprocessing prior to reactor loading.

Abstract

Simulation of Photon Counting Detectors (PCD) is essential to optimize detector design and to assess detector response for photon counting CT. In this work, synchrotron measurements performed with the new PCD tiles, developed by Detection Technology (DT) are compared to simulated data. The detector prototype is a pixelated array of CdTe and CZT pixel with 350 μm pixel pitch. Each pixel has 6 energy thresholds. Measurements were carried out at the European Synchrotron Radiation Facility (ESRF) using pencil beam scan to study charge-sharing and fluorescence escape. The experiment was reproduced in a simulation framework that encompasses physics interaction within the sensor, charge generation, drift, and diffusion and Coulomb repulsion. The comparison focuses on counts of distribution across pixel positions and single pixel spectral response. Results show a good agreement for central pixel irradiation, but discrepancies are observed near pixel boundaries. The assessment of the modeling for interpixel irradiation is currently under study.

Abstract

Radiatio-induced leakage current has been found in ohmic CdTe/CZT detectors under high-flux X-ray irradiation scenario such as photon-counting X-ray CT and new generation light sources. Slowly evolving radiation-induced leakage current that compromises signal stability. In this work, we experimentally characterize the dynamic behavior of this leakage current using step-irradiation and rapid-chopping measurements on a custom 32×32 pixel CdTe detector module. The results reveal a two-timescale response consisting of a fast photocurrent component and a slow, history-dependent leakage component that increases approximately exponentially during irradiation and decays over minutes after beam-off. This behavior is attributed to trap occupancy dynamics that modulate carrier injection at the contacts. A first-order exponential model is proposed to describe the temporal evolution, and a recursive deconvolution framework is explored for signal correction. However, observed nonlinearities and flux/history dependence indicate that a linear time-invariant (LTI) assumption is insufficient, suggesting the need for a more general nonlinear model. These findings provide insight into the instability mechanisms of Ohmic CdTe detectors and offer experimental support for the development of improved dynamic correction strategies for high-flux X-ray imaging applications.

Abstract

CdTe Timepix3 based instruments are an excellent option to perform fine Compton gamma-ray polarimetry. Its high pixelization level (55×55 µm²) allows to maximize instruments modulation factor and to deliver the higher polar angular resolution in the soft gamma-ray domain. This level of performance is especially important for future gamma-ray astrophysics’ telescopes, since high-energy polarimetry is still a domain with a vast research potential. Timepix3 can be operated in two modes: frame mode and data-driven mode. In data-driven mode each interaction full detection data is available: pixel position (x and y), energy and time. In the frame mode only position (x and y) and energy are available for a given time interval, during which all interactions are integrated in a single frame. However, in this mode less intensive front-end electronics’ processing capacity is required leading to substantial lower power consumption. This is an advantage that can be decisive for applications where high sensitive polarimetry must be performed in low electric power platforms, such as cubesat platforms. An experimental set of measurements with a CdTe Timepix3 operating in frame mode was performed at ESRF under a ~99% polarized 200 keV beam. A Q polarization modulation factor of 0.71 ± 0.014 was obtained, proving that under certain conditions and constraints, frame-mode polarimetry is not only a valid method but also a performant one.

Abstract

This work compares the imaging performance of a GaAs photon-counting detector (PCD) and an energy-integrating detector (EID) under various photon-flux conditions. The detectors were characterized using the modulation transfer function (MTF), noise power spectrum (NPS), and detective quantum efficiency (DQE), and were evaluated under similar effective spatial-resolution conditions. An NPS-based noise prediction model was then used to estimate the image standard deviation as a function of photon count. The results show that the GaAs PCD and the resolution-matched EID exhibit comparable frequency-domain performance and image quality under the reference photon-flux condition. At low photon flux, however, the PCD shows a markedly lower predicted noise level. Finally, phantom imaging experiments further confirm that the PCD provides lower background fluctuations and clearer structural visibility under low-flux conditions. These results demonstrate the noise advantage of GaAs photon-counting detectors for low-flux X-ray imaging.