|
Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://xmm.vilspa.esa.es/user/socpub/ccd_review/node14.html
Дата изменения: Fri Aug 30 15:54:52 1996 Дата индексирования: Tue Oct 2 12:14:21 2012 Кодировка: Поисковые слова: п р п р п р п |
CCDs were generally expected to exhibit radiation tolerance comparable with
other CMOS circuits. Early indications from particle physics experiments
indicated graceful degradation at doses of 100krads [Baily et al]. However, the first
experiments with Co-60 irradiation showed loss of charge transfer performance
at doses of 
10krads [Lumb and Holland 1988b]. This was a concern for the
XMM mission, for
example, with a high dose expected during its 10 year mission life (70krads in
thin electronics boxes). These studies showed that the charge transfer
degradation was a more severe problem than conventional integrated circuit
damage mechanisms. Subsequent experiments
with proton irradiation showed that displacement damage creates
measurable loss of charge transfer efficiency
at doses as low as 10's rads, and hence the X-ray applications will require
optimized shielding and CCD architectures to minimize loss of data.
To maximize the tolerance to radiation effects, the CCDs must be made with very narrow charge transfer channels to minimize the interaction with the population of traps. Operation at the lowest possible temperatures also helps to recover response by increasing the trap emission time constants sufficiently to ensure that the traps will tend to remain filled.
A large fraction of the degraded CTE performance can be recovered by annealing
the CCD through raising the temperature to significantly in excess of
100 
C for several hours. Isochronal annealing studies [Holland et al] show that
the main population of traps (but by no means all) generated by lattice
displacement are Phosphorus-Vacancy centers, which anneal at a temperature of
C.
Additional means of limiting the degradation of energy resolution will be employed by directly measuring the degraded CTE values, which may allow the event pulse heights to be reconstructed by applying a pixel position dependent correction to each measured charge packet. (However energy resolution will still be degraded due to the stochastic loss of charge in the trapping centers).
Another manifestation of radiation damage is that of dark current spike
generation. For example, after 
krads of low energy protons, around 60
dark spike pixels were found in an EEV TV format CCD, with -70 C dark
current generation rates of ~5-10 electrons per second [Holland et al]. These would
appear as significant events in each and every image frame. This form of damage
tends to reduce the utility of devices with low dark current obtained through
the techniques of surface inversion (see Section 3.5): bulk dark current sites
start to dominate the dark current signature, and are not affected by inversion.
The use of on-board bad-pixel inhibiting circuits may be employed to remove the
artefacts from the science data, but their effect will extend beyond the bad
pixel to neighboring pixels if event recognition schemes are employed.
These severe radiation damage problems will require careful design of shielding if CCDs are to survive several years in orbit. However for XMM and JET-X the adoption of highly eccentric orbits exacerbates the problem due to the repeated passage through the trapped proton belts. The trade-off here is that the higher orbits allow a greater efficiency of observation time and real-time ground contact, and the use of passive cooling is facilitated, but the effects of radiation damage may seriously curtail the lifetime of the experiment.
Substantial efforts are underway to understand the many aspects of the problem, ranging from the physics of non-ionizing energy loss [Marshall et al], the measurement of damage at different energies and with different device structures, temperature dependency effects and shielding calculations.