The CEI is involved in the mitigation of radiation damage effects to both CCD and CMOS image sensors through the optimisation of device operating parameters and by recommending modifications to the device structure for improved radiation tolerance. We have undertaken a number of radiation damage studies leading to an increased understanding into the post irradiation performance of detectors and specifically the impact of the different defects which are formed as the device is irradiated.
The two main interactions are inelastic collisions with atomic electrons which cause ionisation of the atomic lattice and elastic and inelastic collisions with atomic nuclei which lead to displacement damage, also referred to as ionising and non-ionising damage. The impact of the two damage mechanisms is shown in the Table 1. The severity of each defect is dependent on the radiation type, energy, and fluence, which is determined by the orbital parameter of the spacecraft and its mission duration.
|Damage Mechanism||Location||Resulting Defects|
|Ionisation||SiO2 below the polysilicon gates (gate dielectric)||Generation of a flat band voltage shift, shifting clock and output amplifier bias signals|
|Ionisation||Si-SiO2 interface||Generation of traps at the surface, further increasing flatband voltage shift and surface generated leakage current|
|Ionisation||Region of charge generation and collection||Transient effects|
|Displacement||Silicon where charge is generated, collected, transferred and measured||Decreased charge transfer efficiency in CCDs and Lag in CIS|
|Increase leakage current generation|
|Increased numbers of dark current spikes and leakage current non-uniformity and random telegraph signals|
Table 1: Table showing the radiation damage mechanism and its location of interaction
and the resulting defect
Ionisation occurs when an orbital electron is removed from an atom, via the photoelectric effect, Compton scattering or pair production. Direct ionisation occurs with alphas, betas and protons which are charged, while non-charged radiation, e.g. neutrons, X-rays and gamma-rays, cause indirect ionisation. The net result is the creation of a number of e-h pairs. This number is dependent on the energy required to create an e-h pair, which is independent of the type and energy of the incident ionising particle. The mean ionisation energy required to create an e-h pair is 3.65 eV for silicon and 18 eV for silicon-dioxide, the two materials of relevance for silicon detectors.
The electrons generated within the charge collection and storage regions will be collected in the potential wells, and the holes removed, forming the transient effects, which are read out as signal charge. Pairs generated at the Si-SiO2 interface that do not recombine diffuse through the lattice. The electrons are removed quickly due to their considerably higher mobility. The holes diffuse towards and become trapped at the Si-SiO2 interface of the detector due to the high concentration of impurities, the deeper a hole is trapped within the lattice the longer until it is released as greater energy is required. The increased positive charge results in an increase in the surface generated leakage current, increased clock induced charge and the creation of a flat-band voltage shift.
Displacement is caused when an energetic particle displaces an atomic nucleus from the lattice; this initial displaced atom is referred to as the “primary knock-on” atom. The resulting damage is dependent on a number of variables, including particle type and energy, time and thermal history after irradiation, and impurity levels. The most abundant particle in space is the proton hence these typically form the main source of displacement damage to both CCD and CISs, displacement damage is also caused by electrons with energies above 150 keV, heavy ions, and neutrons. Energetic electrons and protons cause isolated defects, while neutrons, heavy ions, and highly energetic protons produce defect clusters.
Elastic and inelastic collisions can both result in the displacement of atomic nuclei, atoms in non-lattice positions are referred to as interstitials and a vacancy is the absence of an atom from the lattice. The total number of displacements created per incident particle is related to the non-ionising energy loss (NIEL) value (MeV.cm2.g-1). Different defects are mobile at different temperatures, making it important to understand the thermal history of the device after it has been irradiated and ideally perform the irradiation at the mission operating temperature.
The four main stable defects in n-channel are the phosphorus-vacancy (E‑centre), the oxygen‑vacancy (A‑centre) which is the result of impurity atoms in the lattice, and the divacancy (J-centre) consisting of two adjacent vacancies. The origin of the fourth defect is currently unknown. The main stable defects in a p-channel device are the divacancy, the carbon interstitial and the carbon-oxygen interstitial pair, it has been shown that, because of these different trap populations, the p-channel CCDs can offer improved tolerance to radiation induced CTI when compared to their n-channel equivalent.
The emission time constants of the different defects vary with temperature making some operating speeds more sensitive to damage at particular temperatures. Through experimental studies it is possible to probe the different defects that will impact charge transfer, for example performing a ‘trap sweep’ where trap pumping is utilised over a range of transfer speeds to identify the defects created during the irradiation. Optimal timings can then be selected to minimise the amount of charge lost during the charge transfer process at that temperature.