• Lepton Flavor Violation
• MEG
• muon
• rare decay
• trigger
• xenon

The MEG experiment investigates the lepton flavor violating decay μ → eγ with a sensitivity on the branching ratio B(μ → eγ ) of 10−13, 100 times lower than the current limit.
The Standard Model (SM) assumes complete conservation of the leptonic number and then predicts B(μ → eγ )=0 at all the order of the theory. A modified Standard Model to contain the neutrino oscillation phenomena predicts B(μ → eγ )≈ 10−55, an unmeasurable number.
Supersymmetric and grand unification extensions of the SM predict the μ → eγ process to lie around the MEG sensitivity. The experimental search to the LFV in this channel will lead to the first observation on physics beyond the standard model, or set strong constraints on those theories.
The MEG experiment was proposed and accepted by the host laboratory, the Paul Scher- rer Institute in Villigen Switzerland, in 1999. The MEG collaboration developed an innovative technology based on the liquid Xenon scintillation indispensable for this ambitious project. The experiment is completely operative since the 2007, the physics data taking started officially in 2008 and will last until 2012 when the sufficient statistic is supposed be collected.
The MEG collaboration gathers ≈ 60 physicists from 5 different countries, namely: Italy, Japan, Switzerland, United States of America and Russia.
The first chapter of this Phd thesis discusses the theoretical motivations supporting this ex- periment. The principles of the Standard Model are shown, and its modification to contain the neutrino oscillation phenomena. The principal aspects of the SUSY-GUT theories are discussed together with their predictions regarding the μ → eγ decay.
The second chapter describes the state of art of the LFV decay search, in particular in the μ sector, providing also a comparison with the τ one. The second part of the chapter is focused in the μ → eγ search giving an historical introduction and a description of the last experimental result obtained by the MEGA experiment. The chapter ends presenting the experimental requirements for an experiment that aims at 10−13 level of sensitivity.
The third chapter exposes the MEG detector providing a description of the beam line setup, the magnetic spectrometer, the photon detector and the electronic acquisition system. The algorithm reconstruction used in the physics analysis are described in the fourth chapter.
The chapters from the fifth to the ninth talk about the MEG trigger system. Starting from the experimental needs, we show the algorithms developed for the real-time selection. A detailed de- scription of the reconstruction of each observable is given. The custom electronic boards designed
2
for the experiment are presented, particular emphasis is given to the system synchronization tech- nique adopted and the LVDS data transfer between different boards. The firmware implementation of the selection algorithms is described in chapter 7. It illustrates the basic concepts of the code and dedicates particular attention to the use of the Look Up Tables and the live time measure- ment. Chapter 8 and 9 illustrate the procedure developed for the algorithms calibration and the system monitoring. In particular the chapter 9 explain in details the pulse shape discrimination trigger used to select α events in the photon detector starting from the Firmware development, the algorithm calibration and the obtained results.
The last part of this thesis, from chapter 10 to chapter 13, is dedicated to the description of the first physics run taken in the 2008. The tenth chapter describes the physics run, in particular for the detector performance from an hardware point of view.
Chapter 11 presents the online selection efficiencies for the μ → eγ signal. A detailed description of the resolutions obtained for the reconstruction of the online observables is given. Finally a discussion about the global DAQ efficiency as a function of the online selection efficiency and the DAQ live time is given.
Th normalization scheme is described in chapter 12 with a description of the idea behind the normalization scheme. It explains in the details the measurement of all the normalization factors and gives the final result.
The physics analysis procedure, based on a mixture of blind and likelihood analysis is presented in chapter 13. The chapter discusses the principles of the analysis: the definition of the blinding region and the likelihood fit. The probability density functions of the physics observables for the μ → eγ signal, the physics and accidental background extraction are described. The experiment sensitivity for the run has been estimated by using the data just outside the blinded region and simulations methods. Finally the result from the 2008 run is quoted.