Thesis BELLE2-MTHESIS-2017-003

Measurement of the branching fraction of the baryonic decay mode $B^0(\bar{B^0} \rightarrow p p \bar{p} \bar{p}$ at Babar and future prospects at Belle~II

Laura Zani ; Francesco Forti

2016
Department of Physics Enrico Fermi, University of Pisa Pisa

Abstract: The goal of this thesis is the measurement of the branching fraction of the baryonic channel $B^{0}(\bar{B}^{0}) \rightarrow p p \bar{p} \bar{p}$ using the full data set collected by the High Energy Physics experiment BaBar, situated at PEP-II $e^+e^-$ collider, at SLAC National Laboratory (California), and its extrapolation at Belle~II, operating at SuperKEKB at KEK, in Tsukuba (Japan). This thesis also includes the experimental activity performed at the High Technology Laboratories of INFN, Sezione di Pisa, with the Pisa Silicon Vertex Detector (SVD) group of the Belle~II collaboration. In the field of the \textit{beauty} physics, baryonic B meson decays are an important element for better understanding the process of hadronization into baryons. Up to now, the discrepancy between the inclusive branching fraction of all the B meson decay modes with at least a couple of baryons in the final state, measured by ARGUS \footnote{ARGUS was an experiment ran at the electron-positron collider ring DORIS~II at DESY, Hamburg, Germany.} in 1992 to be $6.8\pm 0.6\%$, and the sum of exclusive baryonic channels, averaged on neutral and positive B mesons at less than $1\%$, represents an open issue. The analysis presented in this thesis studies the decays of 471 milions of B meson pairs collected at the $\Upsilon (4S)$ peak during the decade of operation of BaBar. It aims at reconstructing the neutral B meson decays to a pair of protons and a pair of antiprotons. The best candidate reconstruction and the event selection have been optimized by studying the background and signal distributions from official BaBar Monte Carlo samples. Different selection techniques have been tested, based on cuts on the event variable distributions and on multivariate analysis methods. To estimate the branching fraction, I start from the upper limit on the branching fraction for the mode $\bar{B}^{0} \rightarrow \Lambda^{+}_{c} p \bar{p} \bar{p}$ measured at BaBar to be $2.8 \times10^{-6}$ at a $90\%$ confidence level. Scaling the result with the CKM factor $\frac{|V_{ub}|^2}{|V_{cb}|^2}$ due to the $b\rightarrow u$ transition, and including the phase space enlargement factor because of the lower proton mass with respect to the $\Lambda_c$ mass, the branching fraction estimate for the decay mode $B^{0}(\bar{B}^{0}) \rightarrow p p \bar{p} \bar{p}$ is of $10^{-7}$. This channel has a high reconstruction efficiency due to its distinct signature of four charged tracks coming from the same vertex. Despite the limited contribution coming from this mode to the inclusive baryonic branching fraction, it is helpuful for clarifying the strong mechanism of baryon production and for providing experimental support to theoretical models. In the event reconstruction, two kinematic variables are used: the beam energy substituted mass $m_{ES}$ and the energy difference $\Delta E$. The former is the B meson invariant mass where the reconstructed B meson energy is replaced by the beam energy and the latter corresponds to the difference between the B meson energy and the beam energy. These variables quantify the kinematic constraints imposed by the experimental setup and they are used for identifying B meson candidates coming from $\Upsilon (4S)$ decays. For a signal event, the $m_{ES}$ distribution is expected to peak at the B meson invariant mass ($5.279$ GeV/c$^{2}$), while $\Delta E$ peaks at zero. The event selection has been finalized in two phases: a cut-based preselection using kinematic cuts, the particle identification reponse for the tracks detected in the event, and the probability of the common vertex fit, removes most part of combinatorial backgrounds. The suppression of the remaining background is optimized by a Boosted Decision Tree (BDT) selector. It exploits as input variables $\Delta E$, the angular distribution of the B meson candidate in the $\Upsilon (4S)$ rest frame and event shape variables, which are powerful in the rejection of the backgrounds coming from continuum events $e^+e^+-\rightarrow q\bar{q}, \mbox{ } q=u,d,s$, and $e^+e^+-\rightarrow c\bar{c}$, characterized by a \textit{jet-like} behaviour. I compared Monte Carlo (MC) distributions with those obtained from the side band region of on-peak data ($m_{ES}$ $<5.27$) and with the off-peak data collected at 40 MeV below the $\Upsilon (4S)$ energy threshold and therefore not containing any signal events. The extraction of the signal yield consists of an unbinned extended maximum likelihood fit to the $m_{ES}$ distribution of on-peak data, in the whole reconstruction range $5.2 < m_{ES} < 5.3$~GeV/c$^{2}$, after the event selection application. I implemented Monte Carlo studies for modeling the background and signal MC distributions of $m_{ES}$. I set up toy MC studies in order to evaluate the bias on the signal yield introduced by the fitting procedure. From MC studies, the expected significance on the extracted signal yield (10 events with the assumed branching fraction of $10^{-7}$) is around $2\sigma$ and the branching fraction upper limit at a $90\%$ confidence level will be estimated. In thesis thesis, no result from the fit to $\Upsilon (4S)$ on-peak data is going to be presented, since the \textit{unblinding} has not been approved yet. The signal window will remain blind until the end of the analysis, in order to avoid any experimenter's bias. The measurement is mainly statistically limited at the BaBar data integrated luminosity. In the last part of this thesis, the prospects for this measurement at Belle~II will be extrapolated, comparing the BaBar performance with the new detector which is being installed at SuperKEKB factory. A factor 10 improvement in the statistical precision is expected at the Belle~II design integrated luminosity $\int\mathcal{L}dt = 50$~ab$^{-1}$. My involvement in Belle~II experiment concerns also the detector construction activity performed as a member of the SVD Pisa group, regarding the assembling and the electrical testing of the \textit{forward} and \textit{backward} DSSD modules used in the three outer layers of the SVD. The SVD is an essential part of the tracking system at Belle II and its efficiency and resolution determines the quality of the charged tracks reconstruction. I have personally contributed to the assembly and testing of the forward and backward modules performed in the Pisa laboratory.

Note: Presented on 25 10 2016
Note: MSc

The record appears in these collections:
Books, Theses & Reports > Theses > Masters Theses