Radiation damage in CCDs

With the launch of the Galileo spacecraft and the Hubble Space Telescope, astronomical imagery with CCDs from outer space began. Today Cassini, Deep Impact, Chandra, XMM-Newton, and a number of other satellitees and space missions (such as the proposed Constellation-X, DUO, ROSITA, and GAIA space missions) have CCD imagers on-board. With these exciting new windows on the Universe come many unexpected effects in the performance and output noise levels of the CCDs involved. The study of radiation damage in CCDs had occurred in a number of military projects, but the low incident flux and low noise levels needed for astronomy required new laboratory work and the development of techniques to deal with or avoid radiation effects altogether (Cameron et al., 2004; Meidinger et al., 2004a; Meidinger et al., 2004b).

The hostile conditions expected in outer space were not the only radiation source to be concerned about for CCDs. Satellites in low Earth orbit, such as the Hubble Space Telescope, pass through the South Atlantic Anomaly

(SAA) periodically, receiving a healthy dose of high energy protons. The Chandra X-ray observatory's largest factor that reduces observing efficiency is the interruption of observations due to passage through the Earth's radiation belts every 2.6 days. X-ray observations are suspended for ~ 15 hours and the X-ray imager is purposely defocused to minimize damage from low energy (100-200 keV) protons (DePasquale et al, 2004). Solar satellites, such as CAST, are also prone to harsh radiation environments (Kuster et al., 2004). Deep space missions like Galileo and Cassini have a radioisotope thermal electric generator (RTG) to provide power for the spacecraft as well as a neutron dose that bathes the on-board CCD imager. These inherent radiation environments, along with the general space background of cosmic rays and high energy particles from such events as solar flares or planetary magnetic fields, cause both temporary and permanent damage to a CCD in addition to long-term degradation.

Ironically, as CCDs became better astronomical devices in terms of their low read noise and dark currents, they also became much more susceptible to damage by high energy radiation. The SAA, for example, provides about 2000 protons per square centimeter per second with energy of 50-100 MeV, for each passage. Galileo's RTG produced 1010 neutrons per square centimeter at the location of the CCD over the expected six-year mission lifetime. Passage through Jupiter's radiation belts near the moon Io was predicted to provide a 2500 rad dose of radiation to the CCD with each orbit. These levels of radiation do indeed cause damage to the CCD involved and methods of monitoring the changes that occur with time and the development of new manufacturing techniques aimed at radiation hardness were needed (McGrath, 1981).

The two major areas of concern in radiation damage to CCDs are (1) high energy photon interactions, which result in fast electrons, which in turn cause simple, localized damage defects and the generation of numerous electron-hole pairs, and (2) nuclear reactions caused by uncharged neutrons or high energy protons, which cause large area defects and are more likely to lead to partial or complete failure of a device (Janesick, Elliott, & Pool, 1988; Janesick, 2001). The first of these radiation induced concerns is called an ionization effect and involves gamma rays or charged particles. The second, involving massive particles, is termed a bulk effect or displacement damage owing to its ability to displace silicon atoms from their lattice positions within the CCD.

Displacement damage can involve single silicon atoms or bulk damage involving clusters of atoms, all removed from their original lattice locations within the CCD. The vacancies remaining in the lattice structure create trapping locations, which in turn cause degraded or no CTE performance for one or more pixels in the array. As the result of lattice stresses, the trap locations become populated by one or more of the doping elements such as phosphorus. The presence of a phosphorus atom within the silicon lattice modifies the band gap energies locally and is thought to be the cause of observed reduced CTE effects (Srour, Hartmann, & Kitazaki, 1986). The CTI performance of the front illuminated CCDs aboard HST suffered radiation damage from exposure to soft protons when passing through the SAA. The damage increased the CTI by more than two orders of magnitude (Grant et al., 2004) and the observatory team has developed a model of the damage to help mitigate its effect on observations.

Repair of some percentage of single lattice displacement defects (i.e., hot pixels) has been accomplished by cycling the CCD to room temperature or higher and back again to operating temperature, a process called annealing. The back-side, thinned SITe CCDs in the HST Advanced Camera for Surveys (ACS) undergo a routine monthly annealing process. Hot pixels (pixels with enhanced dark current of 0.04 electrons/pixel/s or more), appear at a rate of ~ 1230 per day in the ACS CCDs. Annealing the detectors will fix about 60-80% of new hot pixels (new since the last anneal) but very few of the older hot pixels are repaired. Figure 7.4 illustrates this procedure for the ACS Wide Field Camera (WFC) CCDs in the ACS.

Bulk defects in CCDs are essentially impossible to repair. It has been noticed, however, that at low temperatures (< -100° C), the trapped charge


Fig. 7.4. This figure shows the growth in the number of ACS/WFC hot pixels since installation aboard the HST. One can see the lowering of the hot pixel count through monthly anneals as well as the continued overall evolution to increasing numbers. From Clampin et al., 2002.


Fig. 7.4. This figure shows the growth in the number of ACS/WFC hot pixels since installation aboard the HST. One can see the lowering of the hot pixel count through monthly anneals as well as the continued overall evolution to increasing numbers. From Clampin et al., 2002.

accumulated at the defect location remains trapped and has little effect on the overall CTE. This temperature dependence has been shown to be proportional to exp (-ET/kT), where ET is the activation energy of the lattice traps (Janesick, Elliott, & Pool, 1988). Thus one way to avoid lattice defects is to operate the CCD at temperatures as low as possible. Interestingly, high temperature operation (>30° C) allows trapped charge to be released very quickly, eliminating a deferred charge tail and providing good CTE. These various techniques involving temperature manipulation of a CCD system are often hard to employ practically in space. Additionally, the temporal behavior of the CCD involved can be unpredictable and may be different for each CCD, even those of the same type.

Ionization effects, caused by gamma rays or charged particles, cause a charge buildup in the CCD gate structures and can produce increases in the CCD dark current. Whereas a 150 keV electron is needed to cause an actual silicon atom displacement, only a few eV of energy deposited in the gate insulator is enough to change the potential and cause charge trapping. Even an intense UV flood (Ditsler, 1990; Schaeffer et al., 1990) with 2500 A photons can cause dark current increases or even render the CCD inoperable. The charge buildup causes new states to exist within the band gap of the silicon leading to easier generation of thermal electrons and thereby an increased dark current. The affected CCD pixels, that is, those that have been damaged by the ionizing radiation, will show increased dark current, while their neighbors will not. Histograms of the amount of dark current produced as a function of pixel signal level often show "spikes" of dark current at specific signal levels. This is taken to indicate a sort of quantized structure in the amount of damage that occurs per radiation site (Janesick, Elliott, & Pool, 1988).

Additional damage to CCDs in space by micrometeoroids has recently been studied (Meidinger et al., 2003) for the CCDs on the XMM-Newton X-ray satellite. Other background increasing radiation events occur as well (Freyberg et al., 2004; Katayama et al., 2004) even if their cause remains a mystery.

Methods of protecting CCDs from radiation effects are varied. The type and amount of radiation expected and the scientific goals of the imager must be carefully weighed to produce the final compromise. For example, the CCD flown on the Galileo mission was initially tested for its ability to withstand gamma radiation similar to that expected in the Jovian magnetic fields, a test that it passed well. It was probably by shear luck that a test or two was also performed to understand its performance when exposed to neutrons. Neutron bombardment revealed an increased dark current was prevalent in the CCDs and a redesign of the imager was needed. To mitigate the problem, the CCD imager had its operating temperature changed to -130° C compared with the original specified value of -40° C. In the case of the Hubble Space Telescope CCDs, increased dark current is not a large factor because of their colder operating temperature, but long-term degrading CTE and QE effects have been seen (Holtzman, 1990; Clampin et al., 2002) and attributed to in-orbit radiation damage. A detailed report of the detectors in the Hubble telescope is contained in Brown (1993) and considerations on improving the ability of CCDs to counteract the effects of radiation are discussed in IEEE Nuclear Science Symposium (1988), Bely, Burrows, & Illingworth (1989), and Janesick (2001).

Was this article helpful?

0 0
Telescopes Mastery

Telescopes Mastery

Through this ebook, you are going to learn what you will need to know all about the telescopes that can provide a fun and rewarding hobby for you and your family!

Get My Free Ebook

Post a comment