## Overview of C/A-ratio depth sensing in CZT planar detectors
Depth sensing using the cathode-to-anode (C/A) signal ratio exploits the different weighting of induced charge on the cathode and anode electrodes as a function of interaction depth (z coordinate). For planar CZT detectors (uniform cathode face, segmented anode or pixel/strip anodes) the measured cathode pulse amplitude is approximately proportional to the total induced charge (integrated electron drift plus hole contribution weighted by electrode geometry), while anode amplitudes (especially small-pixel anodes or collecting strips) depend strongly on where electrons are collected and on hole trapping near the anode. The C/A ratio therefore provides a monotonic proxy for interaction depth that can be used for depth estimation, energy correction, and background rejection.
## Basic physical basis and assumptions
The method assumes:
- Single-site energy deposition (or separated multi-site events treated individually).
- Electron-dominated charge collection (CZT: μe ≫ μh) so anode signals largely reflect collected electrons; holes are poorly collected and cause depth-dependent signal deficits.
- Linear or monotonic relationship between true depth z and C/A over the detector thickness (empirically established via calibration).
- Stable biasing and electrode uniformity; effects from lateral position (x,y) are secondary or corrected independently.
## Common C/A-based algorithms
- Empirical ratio mapping:
- Compute C/A = Vcathode / Vanode (peak amplitudes or integrated pulse areas).
- Use a calibration curve C/A ↔ z obtained from collimated measurements or Monte Carlo + charge-transport simulation to map each event to depth z.
- Analytic/semi-empirical fit:
- Fit measured anode vs cathode distributions for monoenergetic photopeaks to a chosen functional form (e.g., exponential + linear terms) that models hole-induced tailing and electron trapping; invert this relation to give z or to compute energy correction factors.
- Example functional forms used in literature: Emeas(C/A) = −a0 exp(−a1·C/A) − a2·C/A + a3, then CF(C/A)=Etrue/Emeas to correct anode energy.
- Look-up table (LUT) correction:
- Build LUT of expected anode signal vs C/A for given energies and map measured events to corrected energy and depth by interpolation.
- Bayesian / maximum-likelihood depth estimation:
- Use forward model (transport + electronics noise) to compute P(C,A | z,E) and estimate posterior P(z|C,A,E); useful when optimizing depth resolution and accounting for noise.
- Multi-parameter fitting with pulse-shape:
- Combine C/A ratio with pulse-rise time or electron drift-time (when available) to improve depth estimation and disambiguate surface events; fit pulses to drift models to retrieve z.
## Practical implementations in planar CZT designs
- Pixelated anode + continuous cathode:
- Standard implementation: measure summed anode pulse (or pixel-specific collected amplitude) and cathode amplitude; derive C/A for each pixel event and apply calibration correction.
- Advantages: simple hardware (single cathode readout) and good z-sensitivity up to full thickness.
- Strip/orthogonal-strip designs:
- Use collecting anode strips and a single or segmented cathode; C/A used per strip to infer depth for each strip-hit; supports tiling with fewer channels.
- Coplanar-grid variants (modified):
- Coplanar-grid collects a differential anode signal to suppress hole contribution; when cathode readout is available, C/A can still provide depth, but coplanar processing changes anode weighting and requires adjusted calibration.
- Drift-strip / focusing-electrode detectors:
- Electrode biasing that focuses electrons to narrow anode collects can sharpen C/A dependence and improve depth precision; calibration must include electrode-specific weighting fields.
## Calibration procedures
- Collimated pencil-beam scans:
- Move a collimated monoenergetic source across the detector edge (or flood irradiate with coincidence tagging) to generate known-depth event sets for mapping C/A to z.
- Coincidence depth tagging:
- Use two detectors in coincidence with known geometry to select interaction planes and derive empirical C/A-vs-depth response.
- Monte Carlo + charge transport modeling:
- Simulate charge generation, trapping, and induction with detector weighting fields to predict C/A behavior; tune transport parameters (μτ for electrons/holes, trapping times) to match measured data.
- Energy-dependent calibration:
- Build calibration curves at multiple energies because trapping/hole effects and pulse formation can be energy-dependent; store energy-binned LUTs or parametric fits.
## Corrections and uses of depth information
- Energy correction (tailing removal):
- Apply multiplicative correction factors CF(C/A) = Etrue / Emeas(C/A) derived from calibration fit to remove depth-dependent low-energy tailing and improve FWHM.
- Depth-dependent event selection:
- Reject events near cathode or anode surfaces that suffer worse charge collection or are likely background (e.g., surface leakage, side-penetrating background).
- 3D event reconstruction:
- Combine C/A depth with x–y from pixel/strip coordinates to localize interactions in 3D for imaging and Compton-sequence ordering.
- Depth-weighted spectral shaping:
- Weight or normalize signals by depth to equalize detector response across thickness for spectroscopy or imaging.
## Limitations and error sources
- Nonlinearity and saturation:
- C/A can saturate near extremes (very close to cathode/anode) reducing depth sensitivity; empirical calibration must handle end effects.
- Multi-site events and charge sharing:
- Overlapping interactions change anode distribution; simple single-site C/A mapping is invalid unless events are separated using multiplicity or pulse-shape.
- Lateral dependence:
- Weighting field variations across pixel/strip can introduce x–y dependence in C/A; residual lateral dependence must be corrected or included in LUTs.
- Electronic noise and baseline variation:
- Noise on cathode (often lower amplitude) degrades depth precision—optimal digitization (pulse shaping, integration windows) is required.
- Temperature, bias drift and polarization:
- CZT properties change with temperature and bias history; periodic recalibration or real-time monitoring improves stability.
- Hole trapping variability:
- Variations in μhτh across crystal lead to event-to-event scatter in the C/A ↔ z relation, limiting depth resolution.
## Performance metrics and achievable resolutions
- Typical depth resolution:
- Literature reports millimeter-scale depth resolution (sub-mm to a few mm) depending on detector thickness, SNR, and electrode geometry; thinner detectors and higher SNR yield better z precision.
- Energy resolution improvement:
- C/A-based depth correction commonly reduces low-energy tailing and improves photopeak FWHM by factors depending on original tailing (examples: from ~6–10% down to ~3–4% at 511 keV in well-calibrated systems).
- Calibration density and stability set ultimate correction accuracy.
## Best-practice implementation steps (concise)
1. Acquire cathode and anode signals with sufficient ADC dynamic range and synchronized timing (sample full pulse shape if possible).
2. Build calibration using collimated sources or coincidence tagging across depths and at multiple energies.
3. Fit empirical model or build LUT mapping C/A → z and Emeas(C/A) → CF(C/A).
4. For each event compute C/A, map to z, apply CF to correct anode energy, and flag events in surface regions for rejection if needed.
5. Monitor operating conditions (bias, temperature) and recalibrate periodically or apply corrections for drift.
## Advanced variants and enhancements
- Combine C/A with drift-time/pulse-rise-time for improved z resolution and to detect multiple interactions.
- Use ASICs with low-noise cathode amplification and high sampling speed to enable pulse-shape fitting and Bayesian depth estimation.
- Implement per-pixel/strip LUTs to remove lateral dependence and achieve uniform spectroscopic response.
CdZnTe Association (CdZnTe.com)
https://www.cdznte.com/blog/what-depth-sensing-methods-using-cathode-anode-signal-ratio-are-applied-to-a-czt-planar-detector.html
