Dark matter search with long-lived particles with the CMS experiment at the LHC

• Ano de defesa: 2025
• Autor(a) principal: Orzari, Breno [UNESP]
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Aceleradores de Partículas; Física além do Modelo Padrão; Matéria escura - Astronomia; Partículas (Física nuclear); Dark matter (Astronomy); Particles (Nuclear Physics); Particle Accelerators
• Link de acesso: https://hdl.handle.net/11449/312795

In this work, two projects are presented: improvements to the CMS pixel tracking algorithm and a search for dark matter particles at the CMS detector. The improvements to the HLT pixel tracks reconstruction algorithm were performed to address two issues: the hard maintainability, given that the reconstruction code is duplicated, and the exclusivity to a given accelerator device in the process of reconstructing events. The usage of a performance portability library allows for the solution of these issues, by having only one piece of the code that is compiled for distinct devices. Besides solving both issues, these improvements will play a fundamental role during the HL-LHC era. The search for dark matter particles is performed in terms of the AMSB model, which contains a charged unstable particle, the chargino, that decays inside the CMS detector into products that are not reconstructed. This process signature in the detector is that of a charged particle track that vanishes in the internal layers. One of the decay products is the neutralino, which is the lightest supersymmetric particle and is a dark matter candidate. The search for disappearing tracks was done using 62.5 fb^-1 data collected by the CMS experiment with a center-of-mass energy of sqrt{s} = 13.6 TeV during 2022 and 2023. Given that the CMS collaboration still blinds this data analysis, the results are interpreted in terms of expected exclusion regions with 95% CLs using data from the control regions of the search. Considering the Run 3 data only, a wino-like chargino is expected to be excluded up to a mass of 777 (405) GeV for a lifetime of 3 (0.2) ns, while a higgsino-like chargino is expected to be excluded up to a mass of 650 (354) GeV for a lifetime of 3 (0.2) ns. When combining this result to the previous Run 2 results, a wino-like chargino is expected to be excluded up to a mass of 904 (511) GeV for a lifetime of 3 (0.2) ns, while a higgsino-like chargino is expected to be excluded up to a mass of 762 (437) GeV for a lifetime of 3 (0.2) ns.