Spectroscopy with pixelated CdZnTe gamma detectors – experiment versus theory

Abstract Methods for obtaining excellent spectroscopy with pixelated CdZnTe gamma detectors are explored experimentally and theoretically. For a monolithic CdZnTe detector with pad segmentation, the dependence of the induced charge on the depth of interaction is found to be much weaker than for a planar detector, even for coarse pad segmentation and despite the large hole trapping. Two methods are described for removing the dependence on depth of gamma interaction for pixelated detectors. One method is by introducing a mild amount of electron trapping to compensate for the almost complete hole trapping found in most CdZnTe detectors. Proper tuning of the electron trapping by adjusting the detector applied voltage can remove the dependence on depth of interaction over most of the detector volume. At the optimum voltage, we obtain an energy resolution of 1.7% full width at half maximum (FWHM) for the 356 keV line of 133 Ba. Excellent quantitative agreement with theoretical calculations provides a basis for detector design based on these calculations. A second method for obtaining excellent spectroscopy is achieved by simultaneously measuring the pad signal and the signal from the common electrode. Individual gamma lines are observed as sharp bands in the correlation profile between these two signals. A correction is performed to linearize these bands so that they will project onto the photopeak. Using this procedure, we obtain for the 356 keV gamma line of 133 Ba an energy resolution of 1.0–1.5% FWHM with a peak-to-valley ratio approaching 100/1.