The55Fe x-ray energy response of mercury cadmium telluride near-infrared detector arrays

The accurate determination of a detector's fundamental parameters, including read noise, dark current, and QE, relies on a proper measurement of a detector's conversion gain (e- ADU-1). Charge coupling effects, such as interpixel capacitance, attenuate photon shot noise and result in an overestimation of conversion gain when implementing the photon transfer technique. An approach involving 55Fe X-rays provides a potentially straightforward measurement of conversion gain by comparing the observed instrumental counts (ADU) to the known charge (e-) liberated by the X-ray. This technique is already preferred within the CCD community, as the pair production energy for silicon is well established. In contrast, to date the pair production energy is unknown for HgCdTe, a material commonly used for near-infrared detectors. In this paper, we derive a preliminary calibration of the 55Fe X-ray energy response of HgCdTe using 8 HST WFC3 1.7 μm flight grade detectors. Our conversion of the X-ray intensities from counts into electrons implements a technique that restores the "true" gain via classical propagation of errors. For these detectors, our analysis yields preliminary results of good statistical precision: each Kα event generates 1849 ± 46 electrons, which corresponds to a pair production energy of 3.21 ± 0.08 eV. We are continuing to assess potential systematic effects to further refine the accuracy of this result.