• Charged Lepton Flavour Violation
• Drift Chamber
• MEG II experiment
• Particle Physics
• Recostruction algorithm

The MEG (Mu to Electron Gamma) experiment was designed for the
search of μ→e +γ decay, a process totally forbidden in the Minimal
Standard Model (MSM) and pratically forbidden also in its extensions: the
observation of this decay would demostrate the existence of the Charged
Lepton Flavour Violation (CLFV) and would be a strong hint in favour of
the theories beyond the Standard Model, which foresee this phenomenon.
The MEG experiment was located at the Paul Scherrer Institut (PSI) near
Zurich (Switzerland) because the PSI provides the most intense continuous
low-energy muon beam in the world, with a positive muon stopping rate
R μ ≈ 10^{8} μ^{+}/s. The experiment collected data from 2009 to 2013 and
published a final result on the Branching Ratio of this decay BR(μ→
e + γ) ≤ 4.2 × 10 ^{−13 }at 90% Confidence Level. This result is the world best
upper limit on BR(μ→e + γ) measured up to now. A major upgrade
of the experiment was projected and is under construction to reach a final
sensivity on BR(μ→e + γ) of 4 × 10 ^{−14} , an order of magnitude better than
the MEG final result. This requires a significant upgrade of all sub-detectors
and a muon stopping rate increase by a factor > 2.

This thesis describes the MEG experiment apparatus and its upgrade
(also called MEG II experiment), with a global description of all detectors
and of their potentialities. The working principle and the global structure of
the MEG upgrade is the same as that of the previous experiment: a Liquid
Xenon (LXe) photon detector, with an improved geometrical acceptance and
a higher granularity, equipped with new photo-sensors (SiPM) instead of
PMTs; a Cylindrical Drift CHamber (CDCH) to be used as positron tracker,
described in a dedicated chapter; and finally a scintillator detector (called
Timing Counter), to measure the positron time. The muon stopping rate is
enlarged from 3 × 10^{7} Hz to 7 × 10^{7} Hz. All these changes require a new
trigger and DAQ electronics, which must handle an almost tripled number
of readout channels with a higher bandwidth.
One of the main topics of this thesis is the description of the CDCH
mounted in the INFN facility in San Piero a Grado, Pisa, of its costruction
and assembling procedures and of its performances. The CDCH detector is
a stereo single volume drift chamber, formed by ten layer of drift cells and
completely filled with low-mass 85:15 Helium-Isobutane gas mixture.

The new detector low mass and higher granularity allows a better positron reso-
lution and higher efficiency with respect to the MEG experiment. The whole
detector construction (as the wiring operation) and the mounting procedures
made in Pisa, together with several tests done on wire layers, are described
in details.
The other main topic of the thesis is the positron recostruction and the
MEG software. The first step of the positron recostruction in the MEG
experiment is the Pattern Recognition (PR) algorithm, a program which
exploits the fast low-level information from the CDCH detector in order to
group crossing points of the particle (also called hits) and then to recostruct
portions of particle trajectory (called track segments). At the end of the PR
stage, the track segments are sent to the Kalman Filter, which has the goal to
fit the whole positron track and to obtain the positron kinematic variables.
In order to further improve the positron recostruction, an intermediate step
is added before sending the track segments to the Kalman Filter: a new
program, personally developed, exploits the φ − z correlation between the
positron hits in the CDCH layers and in the Timing Counter detector, and
associates to a Timing Counter signal one or more track segments in the
CDCH before passing the information to the Kalman Filter. In this way the
Kalman Filter calculation is constrained by the presence of the TC hit at
the end of the positron trajectory (or at least of a portion of it) and track
segments without any spatial correlation with the TC hits can be neglected
since they cannot be part of a signal track.
We are considering also the possible ways to use the combined CDCH-TC
information on a second-level trigger.