Toward fast, low-noise charge-coupled devices for Lynx

Lynx requires large-format x-ray imaging detectors with performance at least as good as the best current-generation devices but with much higher readout rates. We are investigating an advanced charge-coupled device (CCD) detector architecture under development at MIT Lincoln Laboratory for use in the Lynx high-definition x-ray imager and x-ray grating spectrometer instruments. This architecture features a CMOS-compatible detector integrated with parallel CMOS signal processing chains. Fast, low-noise amplifiers and highly parallel signal processing provide the high frame rates required. CMOS-compatibility of the CCD enables low-power charge transfer and signal processing. We report on the performance of CMOS-compatible test CCDs read at pixel rates up to 5.0 Mpixs(-1) (50 times faster than Chandra ACIS CCDs), with transfer clock swings as low as 1.0-V peak-to-peak (power/gate-area comparable to ACIS CCDs at 100 times the parallel transfer rate). We measure read noise of 4.6 electrons RMS at 2.5 MHz and x-ray spectral resolution better than 150-eV full-width at half maximum at 5.9 keV for single-pixel events. We report charge transfer efficiency measurements and demonstrate that buried channel trough implants as narrow as 0.8 mu m are effective in improving charge transfer performance. We find that the charge transfer efficiency of these devices drops significantly as detector temperature is reduced from similar to- 30 degrees C to -60 degrees C. We point out the potential of previously demonstrated curved-detector fabrication technology for simplifying the design of the Lynx high-definition imager. We discuss the expected detector radiation tolerance at these relatively high transfer rates. Finally, we note that the high pixel "aspect ratio" (depletion depth: pixel size approximate to 9:1) of our test devices is similar to that expected for Lynx detectors and discuss implications of this geometry for x-ray performance and noise requirements. (C) The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License.