## Overview of Crystal Defects Impacting Spectroscopic Performance
Crystal defects in
Cadmium Zinc Telluride (CZT) material represent the dominant intrinsic limitation to spectroscopic performance in double-sided orthogonal strip detectors. Spectroscopic quality is quantified by energy resolution (full width at half maximum of photopeaks), photopeak symmetry, peak-to-background ratio, and uniformity of response across the active volume. In strip detectors, where charge must drift over extended distances (5–15 mm thickness) and precise steering fields are required for efficient collection at anode strips, even small defect concentrations produce measurable charge trapping, incomplete collection, and position-dependent variations. These effects manifest as low-energy tailing, broadened photopeaks, and degraded depth-of-interaction estimation from cathode-to-anode signal ratios. While modern traveling-heater-method growth has reduced overall defect densities, residual defects remain the primary factor preventing near-Fano-limited resolution in practical devices, particularly when using discriminator-grade material for cost-effective large-area arrays.
## Tellurium Inclusions and Precipitates as Primary Degraders
Tellurium inclusions and precipitates, typically 1–50 micrometers in diameter and distributed throughout the crystal volume at densities of 10^3–10^5 cm^{-3} in commercial material, constitute the most significant defect class degrading spectroscopic performance. These Te-rich secondary phases form during melt growth due to retrograde solubility and non-stoichiometric conditions. When a gamma-ray interaction occurs near or within an inclusion, the local electric field is distorted, and carriers (especially electrons) become trapped at the inclusion-matrix interface or within the Te phase itself. In strip detectors, this trapping produces incomplete charge collection events that shift into the low-energy tail, broadening photopeak widths by several keV and reducing peak-to-valley ratios. Because inclusions are randomly distributed, they create microscopic non-uniformities that are particularly detrimental in drift-strip configurations, where electrons must traverse multiple drift cells; an inclusion in one cell can cause localized charge loss that cannot be fully corrected by induced-signal algorithms. At higher energies (above 500 keV), where Compton scattering increases the interaction volume, the probability of encountering an inclusion rises, amplifying tailing and limiting overall energy resolution to values worse than 1–2 percent at 662 keV in affected regions.
## Deep-Level Traps and Point Defects Affecting Carrier Lifetime
Deep-level point defects, including cadmium vacancies (V_Cd), tellurium antisites (Te_Cd), and impurity-related centers (such as those involving iron or copper at concentrations of 10^{15}–10^{16} cm^{-3}), act as efficient traps with energy levels 0.2–0.8 eV from the band edges. These traps capture electrons or holes during drift, reducing the effective mobility-lifetime product (μτ) below the ideal 10^{-2} cm²/V for electrons and 10^{-4}–10^{-5} cm²/V for holes. In thick strip detectors, the long drift paths required for high stopping power make even modest trap densities cause cumulative charge loss that varies with interaction depth. This depth dependence distorts cathode-to-anode signal ratios used for DOI sensing and introduces position-dependent energy shifts that cannot be perfectly corrected by lookup tables. Spectroscopic consequences include asymmetric photopeak tailing toward lower energies, increased baseline noise from detrapping events, and reduced ability to resolve closely spaced gamma lines in complex spectra. Because strip detectors rely on precise transverse steering fields, trapped charge also creates local space-charge buildup that perturbs the electric field, further degrading uniformity across adjacent strips.
## Dislocations and Sub-Grain Boundaries Leading to Localized Trapping
Lineage structures, dislocations, and low-angle sub-grain boundaries—often forming networks with densities of 10^4–10^6 cm^{-2}—introduce localized strain fields and dangling bonds that act as shallow-to-midgap trapping sites. These extended defects are more prevalent in larger CZT boules and align preferentially along certain crystallographic directions. In orthogonal strip detectors, a dislocation intersecting a drift cell or strip gap creates a microscopic region of reduced carrier lifetime and altered weighting potential, causing charge clouds passing nearby to experience partial trapping or scattering. This results in event-by-event variability in collected charge that broadens photopeak widths and introduces non-Gaussian tails even after depth correction. For charge-sharing events, which are common near strip boundaries, dislocations near gaps exacerbate charge loss into inter-strip regions, reducing the effectiveness of steering electrodes and limiting sub-strip position resolution. In high-rate applications, accumulated trapped charge along dislocation lines can polarize the local field, producing time-dependent degradation of resolution that is difficult to compensate in real-time processing.
## Zinc Segregation and Compositional Inhomogeneity
Zinc concentration variations arising from segregation during traveling-heater-method growth produce axial and radial gradients in Zn fraction (typically 0.04–0.10 nominal), causing local bandgap fluctuations of several meV and corresponding changes in resistivity and carrier transport parameters. These compositional inhomogeneities span millimeter-scale regions and manifest as systematic position-dependent variations in μτ products and electric-field distribution under fixed bias. In strip detectors, this leads to non-uniform drift velocities across the anode array, distorting transverse steering fields and causing strip-to-strip differences in charge collection efficiency. Spectroscopically, the result is broadened photopeaks when events are summed over the full detector area, as different strips exhibit slightly shifted photopeak centroids. Depth-of-interaction estimation also suffers because the cathode-to-anode ratio mapping becomes position-dependent, introducing errors in trapping correction algorithms. For thick crystals required in high-energy gamma spectroscopy, these gradients amplify along the growth axis, making uniform spectroscopic performance across large-area tiled arrays particularly challenging without individual strip calibration.
## Impurity-Related Complexes and Their Role in Polarization
Trace metallic impurities (e.g., Fe, Cu, Ni) and their complexes with native defects form additional deep traps that promote polarization under prolonged bias or high-flux irradiation. Polarization occurs when trapped holes or ionized centers accumulate space charge that screens the applied field, reducing internal E-field strength over time scales of minutes to hours. In strip detectors, this effect is most pronounced near the cathode plane where hole drift paths are longest, leading to progressive degradation of energy resolution and increased low-energy continuum. Because steering electrodes rely on stable transverse fields, polarized regions cause asymmetric charge steering, increasing charge loss into gaps and degrading position reconstruction accuracy. Polarization-induced non-uniformity is especially detrimental in portable or long-duration spectroscopic applications, where repeated bias cycling or temperature fluctuations can exacerbate trap activation.
## Surface-Related Bulk Defects Propagating from Crystal Growth
Although strictly bulk in origin, defects originating at the crystal surface during boule slicing and polishing—such as subsurface damage layers or propagated Te precipitates—extend several hundred micrometers into the volume and interact with electrode patterning. These defects increase surface-state density and local trapping near the anode plane, where the small-pixel weighting potential is most sensitive. In strip geometries, they elevate inter-strip leakage and reduce the effectiveness of passivation layers, indirectly degrading bulk spectroscopic performance by introducing position-dependent noise and charge loss at strip edges. For shared events, surface defects near gaps amplify cross-talk and incomplete collection, limiting the ultimate energy resolution achievable even after digital correction.
## Combined Effects on Charge Transport and Collection Uniformity
The interplay of these defects collectively reduces the effective carrier lifetime across the detector volume, with the most severe impact occurring in regions of overlapping defect clusters. In strip detectors, where electrons must drift laterally under steering fields over distances comparable to the strip pitch, any local trap reduces the fraction of charge reaching the collecting anode, producing position-dependent collection efficiency maps. Hole trapping, already poor in CZT, is further worsened, amplifying the need for sophisticated DOI-based corrections that cannot fully compensate when defect density exceeds 10^4 cm^{-3}. The net spectroscopic degradation appears as an increase in equivalent noise charge, reduced photopeak efficiency, and non-uniform response across the detector face, all of which are more pronounced in strip architectures than in finely pixelated designs because of the longer average drift paths and reliance on precise field steering.
## Quantitative Impact on Energy Resolution and Position Sensitivity
Defect-induced trapping typically degrades room-temperature energy resolution from a potential sub-0.5 percent FWHM at 662 keV to observed values of 0.8–2 percent or worse, depending on defect density. Photopeak tailing parameters increase by factors of 2–5, directly reducing the ability to resolve complex spectra or perform accurate scatter rejection. Position reconstruction via centroid interpolation suffers from increased variance in amplitude ratios, limiting sub-strip accuracy to 50–100 micrometers rms in defect-prone material versus 20–30 micrometers in high-quality crystals. In imaging applications such as Compton cameras or coded-aperture systems, these effects lower angular resolution and increase background, underscoring why defect engineering remains the central challenge in scaling CZT strip detectors for high-performance gamma-ray spectrometry.
## Mitigation Through Material Selection and Detector Design
While complete elimination of defects is not yet feasible, spectroscopic performance is maximized by selecting low-inclusion-density regions of commercial boules, applying post-growth annealing to dissolve precipitates, and employing advanced passivation and steering-electrode designs to minimize the electrical impact of residual traps. Digital pulse-shape analysis and depth-dependent correction algorithms further compensate for defect-induced variations on an event-by-event basis. In practice, these strategies allow strip detectors fabricated from discriminator-grade CZT to achieve spectrometer-grade resolution, demonstrating that while crystal defects remain the dominant limiter, their effects can be systematically managed through combined material and readout optimization. Ongoing improvements in crystal growth continue to reduce defect densities, promising further gains in spectroscopic performance for future large-area CZT strip systems.
CdZnTe Association (CdZnTe.com)
https://www.cdznte.com/blog/what-crystal-defects-most-significantly-degrade-spectroscopic-performance-in-czt-strip-detectors.html
