Performance in a high-rate environment of triple-GEM detectors for the Phase-2 upgrade of the CMS detector

The Phase-2 upgrade of the CMS experiment includes the installation of a new muon station, ME0, based on triple-GEM technology. ME0 will be positioned in the high pseudorapidity region of the muon system, with background rates up to 150 kHz/cm². The thesis discusses the measurements of the performance of triple-GEM detector stacks for ME0, integrating the electronics in the laboratory, and analyzing data obtained from cosmic rays and test beams with an ME0 stack. The data obtained from the track reconstruction through the stack were analyzed to determine the rate capability in a high-background environment and the time resolution of the stack. The work presented in my thesis is aimed at validating the performance of the prototype of the CMS ME0 detector. In particular, the first prototype consisted of four (out of the six foreseen in the final design) stacked triple-GEM chambers. Data have been collected with the prototype standalone during a test beam campaign at the CERN GIF++ I took part in (in July 2023). Since GIF++ provides both a high-energy muon beam and a radioactive source (^137Cs), the efficiency for muons has been measured by varying the rate of the background photons generated by the source. The trend of the efficiency as a function of the background rate has been used to infer the dead time of the system. In April 2024, I completed the integration of the first final prototype of six-layer ME0 stack. Accordingly, the first validation tests have been carried out both with cosmic rays and in a test beam at the H4 beamline. I determined a procedure for the offline alignment of the chambers within the stack, using data from the test beam campaign abovementioned. Additionally, using cosmic rays, I have performed a measurement of the time resolution of the single chambers for muon hits and of the entire stack for muon segments. In particular, to measure the detector time resolution, I have modified the reconstruction of readout clusters to take into account the presence of bipolar signals due to induction between adjacent strips. In GIF++, with a background particle rate up to about 200 kHz/strip, the efficiency of the four chambers in detecting muons has been measured. The results showed an efficiency drop of ~1-2%, confirming the results obtained in the previous tests performed on the single ME0 triple-GEM chamber. The dead time measured is compatible with the dead time of the frontend electronics (which is about 400 ns per readout channel), if one takes into account the different cluster sizes of muon (~3) and background hits (~1). The average dead time measured is approximately 120 ns, ranging from 67 to 182 ns for the single chambers and partitions. The results obtained have shown a time resolution of the single chambers down to 10 ns, and a time resolution for a six-hit track of approximately 5.25 ns. In CMS ME0 segments will be mostly (~97%) made by 5 or 6 hits. These results are in agreement with the requirement of the CMS Phase-2 upgrade TDR, and are compatible with an excellent BX identification of 97%.