Analysis of On-Orbit ACIS Squeegee Mode Data

The MIT and CXC ACIS teams have explored a number of measures to ameliorate the effects of radiation damage suffered by the ACIS FI CCDs. One of these measures is a novel CCD read-out method called “squeegee mode”. A variety of different implementations of the squeegee mode have now been tested on the I0 CCD. Our results for the fitted FWHM at Al-Kα and Mn-Kα clearly demonstrate that all the squeegee modes provide improved performance in terms of reducing CTI and improving spectral resolution. Our analysis of the detection efficiency shows that the so-called squeegee modes “Vanilla” and “Maximum Observing Efficiency” provide the same detection efficiency as the standard clocking, once the decay in the intensity of the radioactive source has been taken into account. The squeegee modes which utilize the slow parallel transfer (“Maximum Spectral Resolution”, “Maximum Angular Resolution”, and “Sub-Array”) show a significantly lower detection efficiency than the standard clocking. The slow parallel transfer squeegee modes exhibit severe grade migration from flight grade 0 to flight grade 64 and a smaller migration into ASCA g7. The latter effect can explain some of the drop in detection efficiency. There are a few observational penalities to consider in using a squeegee mode. Utilizing any squeegee mode causes a loss of FOV near the aimpoint (4 to 16 strips along the full length of the CCDs), as well as the attendant dead-time increase. Secondly, the cost of the software implementation and its testing will be significant. Lastly, each squeegee mode “flavor” would require lengthy, mode-specific calibration observations. Therefore, since an efficacious, ground-based CTI corrector algorithm is now available (see paper by Plucinsky, Townsley, et al. in this proceedings), a scientific judgment will have to be made to determine which, if any, squeegee modes should be developed and calibrated for use by Chandra observers.