Trial lecture - time and place
Trial lecture: 10:15 am at Lille Fysiske auditorium (V232)
Adjudication committee
- Dr. Pamela Ferrari, Nikhef, Netherlands
- Professor Dieter Röhrich, Institute of Physics and Technology, University of Bergen, Norway
- Professor Andreas Görgen, Department of Physics, University of Oslo, Norway
Chair of defence
Supervisors
Additional information
How does a pixel detector speak? And how does radiation affect the performance of silicon
pixel sensors that will be used in the future phases of the Large Hadron Collider (LHC)?
This Ph.D. thesis aims at answering to these two questions.
The LHC is a huge machine that aims at discovering and studying the building blocks of the
universe and how they interact. To extend its discovery potential, the LHC keeps increasing
its beams’ energy and intensity. To cope with these new conditions, the LHC experiments
are upgrading their detectors. One of the recent upgrades of the ATLAS experiment is the
Insertable B-Layer (IBL), inserted in May 2014. Built with state-of-the-art technologies, this
pixel detector aims at better measuring the trajectory of particles that arise from the ATLAS
interaction region. The IBL sends the collected information as a stream of “words” made of
32 bits: each word is in fact a sequence of '0's and '1's. Some words, the so-called “hit
words”, contain two important pieces of information: the coordinates of the point at which a
particle crossed the detector and the charge collected at this position.
Part of the Ph.D. thesis work was the implementation of an offline software (the byte stream
converter) that decodes these bit words into a format readable by the ATLAS reconstruction
algorithms. Using a metaphor, detectors speak different languages. The “dialect” spoken by
the IBL detector differs from the language used by the other three layers of the pixel
detector. For example, it is able to convey more information in less words by using
“condensed words”: 4 words contain information on up to 10 pixel hits (the same number of
words in the Pixel format only transmits 4 pixel hits). Adopting this feature has the key
advantage of reducing the “detector readout bandwidth”, the number of words sent by the
detector per unit of time. Its usefulness is undeniable in the context of a large number of
words sent by each detector 40-million times per second, as it is the case of the LHC
experiments.
Understanding how a high and prolonged flux of particles can damage a detector is of the
utmost importance for the silicon sensors instrumenting inner tracker at the ATLAS
experiment. Radiation reduces the amount of collected charge by modifying the silicon
structure and thus altering its electrical properties. Different approaches are under study to
counteract this degradation: for example, different materials (e.g. diamond rather than
silicon) or different geometries can be chosen.
The second part of this Ph.D. thesis work focused on the study of the 3D silicon sensor
technology. Contrary to the standard planar technology, where electrodes are implanted on
the surfaces of the silicon wafer, in the 3D layout the electrodes are columns etched through
the thickness of the device. This geometry presents several advantages: among them, it
allows for shorter inter-electrode distance, which in turn leads to better radiation hardness
(e.g. charge carriers have to travel a short distance and are less likely trapped in the
trapping centres caused by radiation damage).
The 3D sensor production process includes a series of masking, etching and doping
procedures.
The devices studied in this thesis were produced by SINTEF, a Norwegian research
institute. Preliminary tests showed poor electrical performance in the vast majority of the
sensors. The origin of these shortcoming was unknown, but suspected to be related to a
faulty step in the production process.
As work for this thesis, the sensors were exposed to increasing levels of irradiation and then
tested under different conditions. Results converged to common features: sensors featuring
good electrical performance show higher charge collection even when exposed to a higher
level of radiation with respect to sensors with worse electrical characteristics. The
performed measurements also highlighted that the possible cause for such poor electrical
performance could be related to the presence of spurious structures in the silicon, caused by
a too-thin masking, that modify the electric field structure inside the silicon bulk. The same
conclusion was independently reached by researchers at SINTEF. Correcting the
shortcoming in the production process allowed SINTEF to produce new sensors that
present very encouraging results and that may be used to instrument the future generation
of pixel detectors in the ATLAS experiment.
Laura Franconi is a Ph.D. candidate in Particle Physics at the University of Oslo. She
obtained her Bachelor’s and Master’s degrees at the University of Bologna. Her work
focuses on the characterisation of silicon detector technologies used in high energy physics
Experiments.