Background rates in several hard X-ray photon counting pixel array detectors

Abstract

Background rates in several pixel array detectors are investigated with an eye toward using them with hard, >6 keV, X-rays in very low-rate experiments - e.g., at signal rates <0.1/s/cm2. Measured background event rates for a detector with an unshielded 0.75 mm thick CdTe sensor on the experimental floor at SPring-8 varied from 0.4/s/cm2 with a 6 keV lower-level discriminator (LLD) threshold to 0.2/s/cm2 with a 75 keV LLD threshold. The background for a detector with a 1 mm thick silicon sensor was smaller, ∼0.08/s/cm2 for a 3keV threshold dropping to ∼0.07/s/cm2 at a 17 keV threshold. These rates are dominated by terrestrial sources, such as gamma rays emitted from trace impurities in concrete, with only a small contribution, <0.01/s/cm2, from direct detection of cosmic ray muons (CRMs). 15 mm of Pb shielding reduces the measured rates to < 0.05/s/cm2 in CdTe and to < 0.02/s/cm2 in 1 mm silicon. Additional processing, “time slicing”, as may be used in low-rate experiments where backgrounds may be most problematic, is shown to reduce the background rates further, to < 0.004/s/cm2 with the silicon sensor, and to between 0.002 and 0.02/s/cm2 for CdTe, where the exact value for CdTe depends sensitively on the detector threshold, and the use of the retriggering and/or the use of dual discriminator thresholds. Background rates are presented as functions of discriminator threshold, shielding, and processing. We also discuss the magnitude of, and the correction for, the limitation of the detector dynamic range that can be introduced by the time slicing. Finally, we present one example where time slicing was used.