Investigation of Properties of Novel Silicon Pixel Assemblies Employing Thin n-in-p Sensors and 3D-Integration

Until the end of the 2020 decade the LHC programme will be defining the high energy frontier of particle physics. During this time, three upgrade steps of the accelerator are currently planned to further increase the luminosity and energy reach. In the course of these upgrades the specifications of several parts of the current LHC detectors will be exceeded. Especially, the innermost tracking detectors are challenged by the increasing track densities and the radiation damage. This thesis focuses on the implications for the ATLAS experiment. Here, around 2021/2, after having collected an integrated luminosity of around 300/fb¹ , the silicon and gas detector components of the inner tracker will reach the end of their lifetime and will need to be replaced to ensure sufficient performance for continued running|especially if the luminosity is raised to about 5x10^35/(cm²s¹ ) as currently planned. An all silicon inner detector is foreseen to be installed. This upgrade demands cost-effective pixel assemblies with a minimal material budget, a larger active area fraction as compared to the current detectors, and a higher granularity. Furthermore, the assemblies must be able to withstand received fluences up to 2x10^16 n_eq/cm². A new pixel assembly concept answering the challenges posed by the high instantaneous luminosities is investigated in this thesis. It employs five novel technologies, namely n-in-p pixel sensors, thin pixel sensors, slim edges with or without implanted sensor sides, and 3D-integration incorporating a new interconnection technology, named Solid Liquid InterDiffusion (SLID) as well as Inter-Chip-Vias (ICVs). n-in-p sensors are cost-effective, since they only need patterned processing on one side. Their performance before and after irradiation is investigated and compared to results obtained with currently used n-in-n sensors. Reducing the thickness of the sensors lowers the amount of multiple scattering within the tracking system and leads to higher charge collection efficiencies after irradiation. Devices with thicknesses between 75 µm and 150 µm are investigated before and after irradiation with different experimental approaches, namely radioactive sources, beam tests, and laser measurements. The obtained results are compared to those gathered for devices using the currently widely used thickness of 285 µm. By implanting the sides of the sensors, the distance between the last active pixel implant and the edge can be considerably reduced, allowing for a compact module concept. In this thesis several steps are discussed to reduce this distance from 1.1 mm down to 50 µm. Subsequently, the performance of the different implementations is investigated. The SLID interconnections offer the possibility to stack sensors and several layers of read-out electronics as well as a reduced minimal pitch and eventually a lower cost. In combination with ICVs it paves the way to 3D-integrated pixel assemblies. These can be further optimised in terms of the active area, thanks to a reduced footprint of the read-out chip. Furthermore, it enables the use of specialised processes for the analogue and digital parts of the read-out chip in the different layers. First assemblies employing SLID interconnections were built and the properties of the interconnection are discussed. Finally, etching of ICVs was started and the present status is reviewed.