CZT Ring-Drift detectors for hard X-ray spectroscopy : an investigation of design by experiment and modelling

• Main Author: Boothman, Victoria H.
• Other Authors: Sellin, P. J. ; Lohstroh, A.
• Published: University of Surrey 2016
• Subjects: 616.07
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.701606

Cadmium Zinc Telluride is increasingly popular as a detector of X-rays at room temperature. The limiting factor in its performance is poor hole mobility causing low-energy 'tailing' on peaks. Ring-Drift is a low-noise, single-carrier-sensing configuration that is the gold standard for spectroscopic silicon X-ray detectors. Combining the advantages of Ring-Drift geometry with the efficiency of CZT could lead to a simple, compact, low-cost system for spectroscopy of hard X-rays with the potential for smaller energy resolution than any such systems currently available. There has been very little research on CZT Ring-Drift devices as yet. In this project, a prototype 3-ring drift detector of 7.5mm diameter x 2.3mm was characterised at room temperature with X- and gamma-ray sources of 60-662keV and by microbeam scanning while the voltages applied to all electrodes were systematically varied. Results showed the crucial influence of the lateral field and its ratio to the bulk field upon the active area, peak position and sensitivity. No hole tailing occurred at 78keV. The maximum active radius extended to 2.3mm, beyond the second ring. The leakage current was very low but energy resolution was limited by preamplifier noise. The CZT material and the device geometry were modelled in 3D with Synopsys Sentaurus TCAD. Line scans were simulated and trends in performance with bias conditions matched experimental data but the model was more severely affected by charge sharing than the real device. The model detector was modified in pursuit of optimum performance. The parameters investigated were ring number, width, pitch and position, combinations of applied voltages and the segmentation of the plane cathode into drift rings. Fields and charge drift were visualised and the active volume was mapped in 3D to improve understanding of the factors governing sensitivity, energy registration and energy resolution. It was concluded that there is no single optimum geometry and bias scheme, but a versatile configuration and a set of different voltage combinations that would perform optimally for different interaction depths.