Innovative readout architecture for Monolithic Active Pixel Sensors

Particle tracking plays an important role in many fields, from the fundamental high energy physics research performed in the experimental halls at CERN, to the latest medical imaging equipment for the treatment of cancer; from industrial cameras used to spot manufacturing defects, to space telescopes studying the universe. In the technology of particle tracking, the leading role belongs to silicon sensors, which over the last decades have improved performance with respect to all key metrics: high spatial and timing resolution, low power consumption, low material budget, decreasing production and exercise costs, and high adaptability to several tracking contexts. In particular, Monolithic Active Pixel Sensors (MAPS), which embed in the same silicon die both the sensor and the signal handling electronics, are the latest incarnation of the silicon sensor paradigm; they show great potentialities to further advance the state-of-the-art in particle tracking. Currently there are many MAPS developments in the worldwide scientific and industrial communities. The ARCADIA collaboration of the National Institute for Nuclear Physics (INFN) aims to develop a MAPS tailored towards large-area and low-power tracking applications. The ARCADIA sensor design is inspired by the ALPIDE sensor, a MAPS developed by the CERN ALICE collaboration which is the current benchmark for monolithic tracking pixel detectors. This thesis first briefly introduces the overall development of ARCADIA sensor, and then it will detail the implementation of the experiment-driven readout, describe validation tests, and report on a preliminary campaign of displacement damage dose effects performed to verify the sensor durability in a harsh environment. Moreover, a very innovative architecture relying on an alternative approach for MAPS readout is discussed, with the intent of developing a sensor suitable for extreme-low power applications (such as for outer space applications). The idea takes advantage of several projections on different axes to reconstruct the addresses of the hits; this approach is made possible by data sparsification in the targeted applications. This architecture is ready at a digital electronics level and its performance has been evaluated both with a mathematical approach, using a specifically developed formalism, and with Monte Carlo simulations reproducing potential applications. Simulations were developed using both a hardware-level language as testbenches for validation, and a software framework capable of providing higher statistics. Preliminary calculations on the power consumption and on the timing performances of the architecture are mentioned. Tentative silicon runs are planned in the near future. The immediate scope of this work is to contribute to the effort to validate the ARCADIA device, currently under development. However the larger scope is an attempt to lay the foundations of new devices capable of improving the state-of-the-art performances in tracking; in this respect, future developments and further analysis will often be presented.