Online program Overview Session: RTSD-09

Abstract

Direct image reconstruction methods for Compton imaging such as Simple Back Projection (SBP) are widely used for their simplicity and robustness. Difficulties arise when imaging low-activity, non-point sources that span a significant portion of the detector’s field of view (FOV). The detector’s directional photopeak efficiency will heavily skew the collected data, and may result in a biased image that does not accurately visualize the extent and intensity of the whole object. This work proposes a SBP imaging algorithm with rotational corrections that smooth out the directional efficiency of the CdZnTe detector and generate a 4π FOV Compton image that accurately visualizes the distribution of an extended radioactive source. Experimental results demonstrate the ability to differentiate between naturally occurring radioactive material (NORM) such as Ra-226 and Th-232 found in the concrete masonry unit (CMU) walls and the lack of NORMs from the windows. This was achieved by placing the detector 6” away from a CMU wall and a floor-to-wall window of roughly the same size and rotating it continuously as it collected data. The resulting Compton image clearly showed regions of high and low intensity corresponding to the locations of the CMU wall and window, successfully mitigating the effects of directional efficiency. Further work will investigate source-to-detector distance dependence and field applications, with the goal of applying similar methods to environmental monitoring and surveying scenarios to visualize areas of elevated radiation levels and contamination.

Abstract

Transmission imaging requires a backing photon-sensing detector to accurately measure the interaction site and energy of each incident photon in future systems for material identification. 3D-CdZnTe detectors currently approximate the centroid of each electron cloud as the gamma interaction site with subpixel position sensing, but at high photon energies this measure may be biased by several millimeters from the true gamma interaction position, creating a noisy image. The ring localization technique constructs a ring of potential locations of the interaction site. The ring has its center at the measured centroid and has a radius equal to the average displacement between the interaction site and the centroid. Overlapping multiple rings on a per-event basis recovers this positional error. To demonstrate this recovery, a Th-228 source was placed behind a tungsten collimator with a 0.5-mm gap, which was aligned across a row of pixel electrodes to create centroid and ring images. Images were made using suspected Compton scattering events with a total event energy > 2 MeV.  Two specific pixels were investigated in both images: one in the exposed region and one in the shielded region. The ring localization image shows an improved signal-to-noise ratio in both the exposed and shielded pixels compared to the centroid image. The pixels in the ring image also demonstrate a statistically significant difference between the average signal in the exposed and shielded regions. Subdividing the pixels reveals a count gradient on the investigated pixel, with the higher average signal in line with the collimator region. Future work will add more event types to get a better measure of the image resolution.

Abstract

Three-dimensional position-sensitive Cadmium-Zinc-Telluride (3D-CZT) detectors are promising in astrophysics. However, the specific space environment and scientific observation requirements impose numerous challenges on the readout electronics, including achieving low noise, high timing precision, and high count rates, while balancing high integration with low power consumption. Thus, we propose MPT3128, a 128-channel mixed-signal application-specific integrated circuit (ASIC) for 3D-CZT detectors. Each channel integrates a low-noise charge-sensitive amplifier (CSA), fast and slow shapers, a 12-bit ADC (shared between two channels), and a time-to-digital converter (TDC) with a resolution of 312.5 ps. Specifically, for cathode timing, it features an adjustable fast-shaper peaking time and a triple-threshold timing scheme. Experimental results demonstrate that the analog front-end achieves low timing jitter (~2 ns even at 4.5 fC charge input), low electronic noise (202 e^- rms), and high energy resolution (5.7 keV FWHM at 511 keV), while featuring fully digital output and control interfaces, facilitating system integration. Furthermore, with all channels active, the triggering rate reaches 55 kHz per channel. The unit has a relatively low power consumption (6.34 mW/channel), making it a promising solution for advanced radiation applications.

Abstract

The growing role of solution-processable materials in novel, sensible radiation detectors has established materials like organic semiconductors and perovskites as promising alternatives to traditional ones. Solution-processable materials posses the unique characteristic to be semiconductors that can be processed in liquid form, and thus are extremely useful to develop thin, large-area and inexpensive devices. Quantum Dots have only recently started to draw attention as candidates for innovative, tunable detectors. These nanoparticles possess wide tunability possibilities, facile synthesis, ambient stability and are constituted often of High-Z elements such as Pb, Cd, Se etc. QDs’ optoelectronic properties are strongly dependent on their size, which is tunable during synthesis. Within a size range for every QD material, it is then possible to select the best ones for each application. Recent results suggest that QD-based detectors have high X-ray sensitivity even with low-power operation, are radiation hard and retain their stability under ambient condition. However, a still unexplored and exciting question is if the best materials for direct radiation detection could still perform efficiently in QD form. A positive answer would mean a new, unprecedented way to fabricate such detectors, exploiting the advantages of solution-processable materials such as low-cost, large-area deposition, casting on flexible substrates etc. With this question in mind, and inspired by the excellent detection characteristics of CdZnTe detectors, in this work we synthesized CZT QD and employed them in drop-cast photodetectors. We assessed their X-ray detection capabilities in the 5-50 keV range with low-power biasing, observing a ready, repeatable response to the radiation. This result, albeit still early, seems to suggest a positive answer to our question, paving the way for the use of already established materials (CZT) in QD form for the development of new, advanced direct detectors.

Abstract

ColorCT is a novel readout ASIC for energy-resolved photon-counting detectors (ER-PCDs) dedicated to spectral X-ray computed tomography (CT). It is designed to sustain very high photon fluxes while preserving spectral performance under strict power consumption constraints. The ColorCT test chip features an 11 × 11 pixel matrix (330 µm pitch) and 8 energy bins. A fast analog front-end delivers short pulses (13 ns peaking time, 24 ns FWHM), enabling high count rate capability. A Digital Reset circuit mitigates pile-up and prevents detector paralysis, ensuring a monotonic increasing behavior of the count rate with the X-ray flux with a saturation value of 30 Mcps. A digital charge sharing correction (CSC) improves the spectral accuracy. The ASIC was coupled to a CdTe crystal to perform X-ray characterizations. Experimental results demonstrate improved count rate and spectral response obtained with the Digital Reset and the CSC, with a power consumption of 3 mW/mm² at 3 10⁷ cps/px (3 10⁸ cps/mm²).

Abstract

CZT Compton cameras have demonstrated great promise for nuclear medicine. We are designing a dual layer CZT Compton camera to enhance angular resolution with respect to current single-layer cameras. An application specific integrated circuit (ASIC) that provide excellent 3D position, energy resolution and high data throughout is required for this. A multi-channel ASIC is presented in this work to overcome these challenges. It is composed of 16 anode readout channels, plus a cathode channel, that each contain a charge sensitive amplifier (CSA), pulse shaper, programmable gain amplifier (PGA), peak detector, and configurable peak detector reset dissipating 12 mW per channel to support low-noise and high dynamic range. The channels are designed for a target energy of 218keV & 440keV used in ²²⁵ Ac Compton imaging, with a worst-case 0.81% non-linearity over the dynamic range of 30 keV to 2.2 MeV. Each channel can handle up to 2 nA of leakage current without performance degradation and yields 1.33% FWHM at 662 keV with a single-pixel capacitance of 29.1 pF for each (1.225 mm)x(1.225 mm) pixel. The rise time of each CSA is 150 ns while the shaper peaking time is 320 ns to enable fast readout with a configurable gain of 7, 13, 45, 81, and 132 mV/fC from the PGA. This work demonstrates a compact, scalable readout platform for CZT-based Compton imagers and other room-temperature semiconductor detector systems requiring high energy resolution.

Abstract

CdZnTe (CZT) has been the most commercially successful semiconductor material for room-temperature X-ray and gamma-ray radiation detector applications. It is considered as the gold standard for detector applications that require spectral and often spatial information in a compact instrument. However, CZT still contains a high density of performance-limiting defects such as high concentrations of sub-grain boundary networks, Te inclusions and compositional inhomogeneity in as-grown ingots. Adding a small amount of selenium (Se) in the CZT matrix was found to successfully mitigate the performance-limiting defects and provide for better compositional homogeneity. Thus, the new quaternary Cd_xZn_1−xTe_ySe_1−y (CZTS) is emerging as a next-generation compound semiconductor, potentially surpassing the material quality of CZT. However, the purity of CZTS material is not on par with commercial CZT, principally due to the lack of availability of commercial high-purity CdSe. The presence of the impurities in CZTS caused by the lower-purity CdSe starting material was found to impose a detrimental effect on the charge transport properties of CZTS, resulting in adverse effects on the charge collection and detector performance. Here, we report for the first time on the growth and characterization of Cd_0.9Zn_0.1Te_0.98Se_0.02 grown by vertical Bridgman and by the traveling heater method (THM) using higher-purity CdSe starting material.

Abstract

Directional solidification (DSM) enables rapid growth of cadmium zinc telluride (CZT) ingots with superior scalability and geometric flexibility compared to the travelling heater method (THM). However, the strongly non-equilibrium conditions inherent to DSM promote gas bubble nucleation, growth, transport, and coalescence, leading to porosity that degrades detector performance. Bubble-induced defects distort internal electric fields, enhance charge trapping, reduce the carrier mobility–lifetime product ($\mu\tau$), and increase leakage current and electronic noise through impurity segregation and secondary phase formation. In contrast, THM operates closer to thermodynamic equilibrium and therefore minimizes such defects, albeit at reduced growth rate and scalability.

DSM-grown CZT typically exhibits axial heterogeneity, with increasing bubble density and size toward the last-to-solidify region. This results from coupled interactions between volatile-species thermodynamics, buoyancy-driven transport, capillarity, coalescence, and solidification kinetics. Bubble coalescence and accumulation in the upper melt region significantly amplify porosity near the end of solidification, producing larger pores that are more easily entrapped and more detrimental to detector uniformity.

To describe these mechanisms, a physics-based dimensionless framework is introduced using transport, thermodynamic, and interfacial scaling parameters, including an escape number comparing bubble migration and interface growth velocities, as well as Bond and Capillary numbers governing deformation and stability. Within an atomistic-informed thermophysical model at 5N Plus, temperature- and composition-dependent properties are used to evaluate these parameters and define a predictive merit function for DSM optimization. This framework identifies conditions that promote bubble escape, suppress coalescence and entrapment, and reduce porosity while preserving the advantages of rapid CZT crystal growth.