Disruptive Technologies for Electron Bombarded Active Pixel Sensors

A variety of sensor types have been operated in electron bombarded mode, including CCDs CMOS sensors, and silicon sensors (pixellated photodiodes) in conjunction with active pixel sensors (e.g. Medipix [2]). This project aims to develop a photon counting capability for the TDCpix [3], a newly developed pixel sensor with exceptional timing resolution. It follows on from a previous PIPSS and BBSRC-funded collaboration with CERN to develop a multi-channel photon-counting detectors with picosecond event timing for life science applications. Our original IPS project utilized a microchannel plate detector with CERN-developed preamplifier and time-to-digital ASICs. The recent development of the TDCpix active pixel sensor by the same group at CERN offers comparable time resolution
(100 ps binning, and electronic resolution ~30ps) but with a much higher pixel count (40 x 45 pixels, 12 x 13.5mm^2)), a much higher level of miniaturization provided by integration of the entire electronics on to the chip, and a greatly increased overall count rate capability of ~130 Mcount/s per ASIC, an order of magnitude higher per unit area than its microchannel plate based predecessor.

An electron bombarded TDCpix would offer unrivalled performance with commercial potential for applications using time-correlated single photon counting (TCSPC) such as high content cell screening and other expanding fields in the life science sector, LIDAR instruments for remote sensing, and a variety of other event timing applications where only small arrays of individual photomultiplier tubes are the norm.


Our aim in this project is to identify and develop a technology for photon counting detectors using electron bombarded silicon devices, in order to remove the active pixel sensor from within the vacuum tube, thus greatly simplifying design, de-risking the manufacturing process, and enhancing performance. Removing the chip from the tube will eliminate undesirable elements such as high density vacuum electrical feedthroughs, materials with poor vacuum compatibility, and internal bump, wire, and chip bonding, and will lift the restrictions imposed by these on tube processing which impact manufacturing yield, device reliability, and ultimately, sensor lifetime. Given a successful outcome to this project, we intend to propose a follow-on IPS project, one of whose goals would be to incorporate an additional, relatively low (x20) gain stage using a linear mode electron avalanche process within each pixel of the silicon sensor, matched to the requirements of electron bombarded operation. This will allow the electron bombardment gain to be lowered, reducing the tube operating voltage to safer levels, and reducing the lifetime-threatening radiation damage.

The other elements of an electron bombarded detector design, the vacuum tube including photocathode, and the silicon sensor, will be provided by our industrial collaborators; Photek Ltd., and Micron Semiconductor Ltd, respectively. Photek have extensive experience of design and manufacture of custom vacuum-based detectors with specific expertise in the electron bombarded mode devices, having manufactured an electron bombarded Medipix-based detector. Micron Semiconductor have substantial experience and heritage producing large quantities of custom pixellated silicon sensors for harsh radiation environments at CERN LHC and other similar experiments. Specifically for this project, they have developed a thin entrance window technology which is highly desirable for electron bombarded mode to minimize photoelectron energy loss. The thickness of their currently available Type-9.5 window is 500 Angstroms, and a Type-10 window is under development with a thickness goal of 200 Angstroms. Micron also have a bump-bonding capability necessary for the interconnect development.