Holmkvist, William and Muenstermann, Daniel (2022) Investigation of Post-Processing Resistivity Modification Methods for HV-CMOS Pixel Detectors. PhD thesis, Lancaster University.
Abstract
Silicon detectors plays a key-role in the measurement of particle trajectories in cur- rent particle collision experiments. Future colliding experiments will impose even larger demands on the performance of the detectors. To meet these demands, while simulta- neously keeping costs low, monolithic active silicon sensors using HV-CMOS technology are currently investigated. While showing good promise, many semiconductor foundries do not allow for custom substrate resistivity, making it interesting to investigate whether standard substrate HV-CMOS sensors are a viable option for future high energy particle detectors. To this end, this thesis presents a study on how the charge collection and effective doping concentration are affected by hadron irradiation in HV-CMOS detector prototypes. The investigated prototypes were produced at two different foundries, Austria Micro Systems and LFoundry, and were irradiated with either protons or reactor neutrons. Edge-TCT, using pulsed IR light injection, was the main tool of investigation. Proton irradiated samples were found to perform significantly better after fluences of 5 · 1014 neq/cm2. The neutron irradiated samples had a significant reduction charge collection in spite of an increased depletion region after 1015 neq/cm2. Additionally, all samples showed a decrease in charge collection in the fluence region 1013 - 1014 neq/cm2. Following these, low-resistivity HV-CMOS detectors might not be optimal for some particle collider environments, but could be of use for prototyping with foundries not offering custom resistivity substrates. Additionally, as an attempt to improve performance Thermal Donor introduction was investigated as a method to increase the active region in low resistivity HV-CMOS sensors. Samples were annealed at 450 ◦C at increasing time intervals, and measured using edge-TCT. Introduction rates were found to deviate greatly from estimated values, and Thermal Donor introduction was ultimately concluded not viable as a method to improve detector performance.