1. The onboard calibration for the spaced-row charge injection of the Suzaku XIS Hideki Uchiyama, Yoshiaki Hyodo, Hiroya Yamaguchi, Hideyuki Mori, Takeshi Go Tsuru, Hironori Matsumoto, Katsuji Koyama (Kyoto Univ.), Ken'ichi Torii, Satoru Katsuda, Hiroshi Nakajima, Kazuto Hasuike, Kiyoshi Hayashida, Hiroshi Tsunemi (Osaka Univ.), Hiroshi Murakami, Tadayasu Dotani (ISAS/JAXA), Gregory Prigozhin, Steve Kissel, Eric Miller, Beverly LaMarr, Mark Bautz (MIT) and Suzaku XIS team

2. X-ray astronomy satellite Suzaku & X-ray Imaging Spectrometer (XIS) Suzaku The 5th Japanese X-ray satellite X-ray Imaging Spectrometer (XIS)  X-ray CCD camera onboard Suzaku  3 Front Illuminated (FI) & 1 Back Illuminated (BI) sensors  Positional resolution  ~2’ with X-ray telescope (HPD)  High energy resolution ~140eV@5.9keV (FWHM) in Aug. 2005 (just after launch)  Equipped with a charge injection (CI) structure CCD chip of XIS Suzaku satellite

3. Degradation of the energy resolution Time since 2000-1-1 00:00:00 (s) May 2006 150 200 Aug. 2005 just after launch Energy resolution (eV,FWHM) 1.8×10 8 1.9×10 8 2×10 8 The increase of charge transfer inefficiency (CTI) causes this degradation. The time dependence of the energy resolution @5.9keV

4. Transfer direction Q Q’ Q’’ ACTY Readout node Q’’ 1026 pixel Q’ Charge trap caused by radiation damage in orbit Readout charge 0 Energy resolution is degraded. Q Q X-ray Imaging area Frame-store region ACTY ACTX How the CTI causes the degradation

5. Transfer direction Readout node Q’’ 1026 pixel Q’ Spaced-row Charge Injection (SCI) Charge trap caused by radiation damage in orbit 54 rows Injected charge Q Readout charge 0 Q’Q’’ Energy resolution is recovered! ACTY Injected charge fills the traps. →It decreases the CTI. Q Q X-ray Imaging area Frame-store region Charge injection structure ACTY ACTX

6. We operated XIS with the SCI technique in orbit for the first time in August 2006 and verified the recovery of the energy resolution from 210eV to 150eV (@5.9keV,FWHM) We studied the gain non-uniformity in the SCI mode and developed a new correction method for it. Spectra of onboard cal source 55 Fe Mn I Kα Mn I Kβ Pulse height (channel) no-SCI SCI

7. How the SCI mitigates the gain non-uniformity Perseus cluster of galaxies  It is a bright diffuse X-ray source.  It covers almost all of the FOV of the XIS.  Its spectrum has strong He-like Fe Kα line and the line center is roughly independent of the position. →Suitable for the study of the gain non-uniformity over the chip.  Aug. 2006, in the both SCI & no-SCI mode, exposure time ~50 kilo seconds XIS image of Perseus cluster

8. How the SCI mitigates the gain non-uniformity Perseus cluster Pulse Height channel He-like Fe K α ACTX ACTY cnts/ch/s Pulse Height channel The CTI became ~30% by the SCI. He-like Fe Kα from Perseus cluster

9. Ground experiment of the SCI with proton-damaged CCD 350 400 450 500 100200300400500 Tomida et al. 1997 Positional dependence of Mn I Kα line Transfer pixel “Saw-tooth” gain non-uniformity appeared. → Does it appear in orbit? The gain non-uniformity on a smaller spatial scale Pulse Height Charge Injected row Transfer direction

10.  They irradiate the upper edge of XIS continuously. → We can use long exposure data.  The spectrum has Mn I Kα line and the line center is 5.895keV.  April 2007, in the SCI mode, exposure time ~1 mega seconds XIS image of Cal source Onboard calibration sources 55 Fe

11. A C B 54rows Transfer direction Charge injected row Pulse Height channel Transfer direction A B C ACTY Charge injected row PH CH Mn I Kα Kβ cnts/ch/s Mn I Kα from onboard cal source “Saw-tooth” gain non-uniformity appeared in orbit. The gain non-uniformity on a smaller spatial scale  Onboard calibration source 55 Fe

12. Correction for the gain non-uniformity We modeled the positional non-uniformity of the gain by “saw-tooth” function below. Three parameters, the slope at ΔACTY=0, ΔACTY=54 and PH 0 specify this function uniquely. Fit the observed ACTY-PH relation with the saw-tooth function to obtain the the slope at ΔACTY=0, 54 and PH 0, then correct the observed PH. ACTY ΔACTY PH 0 54 rows charge-injected rows The slope at ΔACTY=54 The slope at ΔACTY=0 PH Saw-tooth function Saw-tooth Correction

13. Result of the saw-tooth correction method Pulse Height channel Without Saw-tooth correction fitting result of the saw-tooth function With Saw-tooth correction Mn I Kα from onboard cal source The correction reduced the gain non-uniformity at 5.9 keV from ~ 0.5% to ~ 0.1% →The gain is uniform on the order of a few eV at 6 keV with this correction. The energy resolution is also improved. 151 eV → 148 eV @5.9 keV FWHM (additional width ~ 30 eV)

14. The current status of the Suzaku XIS software & data processing for the SCI The saw-tooth correction has been built into public Suzaku analysis software and released. Version 2.0 data processing incorporates the saw-tooth correction and the calibration of the SCI. The processing has already begun. The calibration database for the SCI is being updated continuously now.

15. Summary Suzaku XIS operated the spaced-row charge injection technique for the first time and verified the recovery of the energy resolution. The CTI became ~30% by the SCI. The gain non- uniformity over the chip was mitigated well in orbit. The saw-tooth gain non-uniformity was shown in orbit in April 2007. We developed the saw-tooth correction method. It made the gain in the SCI mode uniform on the order of a few eV at 6keV. The software for the SCI mode has been released. The data processing for the SCI mode already started.