Search for rare boson decays with the CMS detector at LHC and the CMS Level-1 trigger Data Scouting

This thesis presents the search for the rare decays of the Z and Higgs bosons to a photon and a charmonium state $\text{J}/\psi$ or $\psi\text{(2S)}$, which subsequently decays to a pair of muons. The channel under study is a probe of the Higgs coupling to the charm quark, thanks to a predicted direct decay process going through a charm quark loop, and it is characterized by a low amount of standard model backgrounds. The similar Z decay is included in the study as the sensitivity needed for an observation is expected to be reached first, offering a benchmark for the validation of the theoretical prediction of the branching fractions. New and unexpected physics might enter in the loop processes and affect the observed decay rates, making the study of rare Higgs boson decays to a meson and a photon a particularly promising ground for anomaly searches. The dataset for the study corresponds to an integrated luminosity of $123 \ \text{fb}^{-1}$ of proton-proton collisions at a center-of-mass energy $\sqrt{s} = 13 \ \text{TeV}$ collected during 2016-2018 with the CMS detector at the LHC. No evidence for branching fractions of these rare decay channels larger than predicted in the standard model is observed. Upper limits at 95\% confidence level are set: $\mathcal{B}(\text{H}\to\text{J}/\psi\gamma) < 2.6 \times 10^{-4}$, $\mathcal{B}(\text{H}\to\psi\text{(2S)}\gamma) < 9.9 \times 10^{-4}$, $\mathcal{B}(\text{Z}\to\text{J}/\psi\gamma) < 0.6 \times 10^{-6}$, and $\mathcal{B}(\text{Z}\to\psi\text{(2S)}\gamma) < 1.3 \times 10^{-6}$. The ratio of the Higgs boson coupling modifiers $\kappa_{c}/\kappa_{\gamma}$ is constrained to be in the interval $(-157,\, +199)$ at $95\%$ confidence level. Assuming $\kappa_{\gamma}=1$, this interval becomes $(-166,\, +208)$. The analysis is statistically limited and it would profit from higher trigger efficiencies. This limitation is more generally affecting the class of Higgs boson decays to a meson and a photon and to a pair of mesons, particularly when the mesons involved are the \( \rho \) and the \( \phi \). This thesis presents a novel data taking and real-time analysis approach for the CMS Phase-2 upgrade at High-Luminosity LHC that allow to perform the analysis using the Level-1 trigger (L1T) reconstruction and to bypass the event rejection imposed by the L1T selection. The system implementing this idea, known as Level-1 trigger Data Scouting (L1DS), will collect and perform an online analysis at the full event collision rate on the reconstructed Phase-2 L1T objects, which will have efficiency and resolution often comparable to the offline one. The L1DS system will enable the search for signatures that feature a too-large irreducible background and therefore currently rejected by the tight requirement on the L1T output rate budget, and signatures for which the background rejection requires algorithms that do not fit within L1T latency and hardware resource constraints. Additionally, it will provide vast amount of data for fast detector diagnostic. A demonstrator of the L1DS system, assembled and commissioned in 2023, is presented. Its architecture and results after its successful deployment in production starting from 2024 are also discussed. The demonstrator will allow the test of different approaches to acquire and process online data from the L1T system, thus shaping the development of the Phase-2 L1DS on real collision data. Projections of the L1DS performance on the rare Higgs boson decays to a meson and a photon or to a pair of mesons are carried out on simulated data. The results highlight how the scouting approach is suitable for the search of these elusive decays, which might open the path to new measurements of light quark flavor couplings.