Which part of a CCD pixel is sensitive to the proton damage?

Abstract

A mesh experiment is, so far, the only practical technique to study charge-coupled devices (CCD) response with subpixel resolution. Since the Chandra ACIS was seriously damaged by low-energy protons in orbit, we undertook a mesh experiment with a relatively low-energy proton beam. The CCD used was designed for the Solid Slit Camera (SSC) employed by the Monitor of All-sky X-ray Image (MAXI) mission that is scheduled to be onboard the International Space Station in 2007. It consists of 1024×1024 pixels, each 24 μm square. Each pixel is equipped with a ‘notch’ structure in the center to make it radiation hard. A proton beam energy has been selected so that the protons passing through the mesh holes, 2 μm in diameter, will penetrate into the CCD chip and damage the transfer channel. At the beginning of the experiment, we used a low intensity proton beam and operated the CCD in proton count mode in order to precisely determine the mutual alignment between the mesh and the CCD. Then, we irradiated the CCD with a strong proton beam, which increased the charge transfer inefficiency (CTI) from 10 −6 to 10 −4. After the proton irradiation, we removed the mesh and measured the CTI with X-rays. In this way, we determined the CTI for individual pixels. We see that there is a high CTI region running through the center of the pixels from top to bottom. This corresponds to the notch region where the charge is transferred. Pixels whose notch structures are damaged show a CTI of ∼3×10 −4 that is about three times worse than those damaged outside the notch regions. When the proton damages in the notch structure, the signal charges are easily trapped, resulting in a high CTI. Our results clearly show that the non-uniformity of the CTI is due to the notch structure. This result is consistent with other experiments showing the effectiveness of the notch structure.