Study of the stitching region and $\eta$-correction of novel passive CMOS sensors for the HL-LHC upgrade of the CMS pixel detector

Passive CMOS silicon pixel sensors are a novel sensor type for pixel detectors in HEP and are currently a candidate for the Phase-II upgrade of the CMS detector. Passive CMOS sensors produced in a commercial 150 nm CMOS process use stitch- ing for the production of large scale sensors. A study of the impact of stitching on the performance of unirradiated sensors is presented in this thesis. Testbeam data with 120 GeV pions has been analysed and the measured cluster size, collected charge, diffused charge and detection efficiency in pixels with stitching has been compared to pixels without stitching. No difference between pixels with and without stitch- ing was measured and no influence of the stitching on sensor performance could be observed. An additional study of the sensor resolution for two pixel clusters has been performed. Non-linear charge division worsens the spatial resolution of pixel detectors. The $\eta$-algorithm is able to efficiently correct for this effect. The impact of this algorithm on the resolution of two pixel clusters for passive CMOS sensors has been investigated. The $\eta$-correction significantly improves the spatial resolution by a factor of two for angles below 4°. No improvement is seen anymore for 9° as expected for the pixel dimensions.