Charge Sharing in Single-Photon-Counting Detectors

Semiconductor pixelated position sensitive sensors are widely used in imaging and spectroscopic applications. The integration of analog and digital structures and CMOS-technology-scaling made implementation of detectors with pixel sizes even down to tens of micrometers possible. A hybrid pixel detector consists of a pixelated sensor connected electrically and mechanically via bump-bonding to an array of electronic readout channels. When an X-ray photon interacts with a detector, it deposits energy in a sensor volume. The electron-hole pairs are created in the sensor, and the charge carriers drift towards the electrodes as a result of a bias voltage applied. Detectors can be divided into those working in the integrating and single-photon-counting modes. In the integrating systems, signals induced by the incoming photons are integrated over a given time exposure, amplified, and digitised. However, this mode of operation, proposed to deal with high intensity of photon flux, has a limited dynamic range and a low signal-to-noise ratio because of integration of the noise. In the single-photon-counting (SPC) systems the photons are processed one by one, and in general, both the photon energy and the interaction position can be measured. Such a solution offers noise reduction, high dynamic range, and enhanced spatial resolution. However, SPC systems face limitations in operation under high-flux conditions. The answer to this problem is, on the one hand, optimisation of the readout channel processing speed and, on the other hand, the minimisation of the pixel size, which allows a detector to accept more photons per unit area. However, when the pixel sizes become smaller, the charge sharing phenomenon starts playing an important role. While charge clouds drift to the electrodes, they spread due to diffusion and repulsion. An increase in the cloud size may result in charge collection by several neighbouring electrodes, and this phenomenon is called charge sharing. In this case, with no circuitry dealing with charge sharing, a detector may register more than one simultaneous event, and the signal amplitudes are no longer proportional to the deposited energy. Therefore, charge sharing may significantly distort the energy and spatial resolution, as well as result in an increase in the false counts within lower energies in the spectrum and a decrease in the number of counts in the higher energies in the spectrum. To overcome problems caused by charge sharing effect, hardware on-chip algorithms have been proposed, and some of them were implemented in silicon. In some applications, an off-chip correction of measured data is also possible. The solutions involve the process of total charge estimation based on fractional signals and assigning a hit to the pixel with the largest charge deposition. Alternatively, the pattern recognition technique or searching the centre of charge-cloud gravity has been proposed. Performing these tasks always requires the development of a large interpixel communication system. Interestingly, even though the charge sharing effect has a negative impact on the spatial resolution of a detector, the resolution can be improved beyond the resolution determined by the physical pixel size when information on the proportions of charge collected by neighbouring pixels is used to estimate the position of the event. This chapter presents the physics of the charge sharing effect, discusses different implementations that attempt to overcome charge sharing, and possible ways of using the charge sharing effect to the advantage of the detectors.