The radiation damage is the main limitation for the operation of position sensitive
silicon detectors at future high energy physics experiments, which aim for ever larger
energies of colliding particles with larger energies. Silicon detectors have been widely
used in all experiments over the last decades. It was however expected, that the
signal from planar silicon detectors would degrade with irradiation to a level where
the efficient operation of the innermost tracking detectors at the upgraded LHC
experiments would become impossible. However, recent measurements with planar
detectors where n+ side is segmented for readout (n+‐p or n+‐n detectors) showed a
charge collection efficiency sufficient for the efficient operation even at the highest
fluences expected at the HL‐LHC, in excess of 1.6∙1016 hadrons/cm2. The key
condition was the operation at very high bias voltages of around 1000 V. Several
groups reported charge collection efficiencies much larger than expected, in some
operating conditions even exceeding the one before irradiation. This is a clear
evidence for charge multiplication in silicon detector, due to impact ionization.
As charge multiplication may well be the reason for successful operation of heavily
irradiated silicon detector, the main focus of this thesis is on this phenomenon, as
well as understanding the device model and operation of heavily irradiated silicon
detectors. Both planar and so called 3D devices of different thicknesses (75, 150,
300 μm), coming from different manufacturers (HPK, Micron, MPI‐HLL,
Soitec/MPI‐HLL, CNM) and irradiated with different reactor neutrons and 200 MeV
pions (and a combination of both) up to fluences of 1016 cm‐2, were investigated
using different detector characterization techniques: Edge‐Transient Current
Technique (Edge‐TCT) and multichannel readout of induced charge by custom made
ASICs.
Edge‐TCT is a novel technique utilizing an short pulses (~ 100 ps) of infra red light
(1060 nm) directed at a polished edge of the detector. Electron hole pairs generated
along the narrow beam (spot size FWHM < 10 μm) are separated by electric field in
the detector and consequently induce currents in the electrodes. As the position of
the beam is externally controlled by moving stages the profiling of the electric field at
different depths is possible in accurate way. The analysis doesn’t depend on time
evolution of the induced current pulse hence the precise knowledge of effective
trapping times is not required for determination the drift velocity, charge collection
and electric field profiles in heavily irradiated silicon detectors.
The Edge‐TCT measurements of the induced current gave first direct observations of
charge multiplication in heavily irradiated silicon strip detectors, taking place in high
electric fields near the main junction (strips). The amplification was found to increase
with detector post‐irradiation annealing, which in this work was studied up to
40960 min at 60 ⁰C. Long term annealing causes build up of negative space charge at
the n+‐p junction, consequently resulting in very high electric fields, sufficient for
initiating impact ionization. A strong correlation between the increase of charge
collection and the increase of the leakage current was also found. These findings
were also confirmed by charge collection measurements with 90Sr electrons.
TCT measurements where detector surface was illuminated were also performed on
special types of miniature detectors, with junction implants not fully covered by
metal, allowing proper analysis of charge multiplication at implant edges, where it
was confirmed to be the highest. Charge sharing between electrodes due to trapping
(incomplete carrier drift) was also studied.
According to the obtained results, an appropriate modeling of the electric field in
irradiated detectors was proposed. A simple model, assuming two space charge
regions at each side of the detector and neutral bulk in‐between was found to
describe the field profile in neutron irradiated detectors. Pion‐irradiated detectors
were found to have strikingly different profiles and attributed to large oxygen
concentration in the detector bulk. The model parameters were also studied during
long term annealing and it was found that the space charge near the main junction
shrinks, which leads to these high electric fields and consequently impact ionization.
The model parameters extracted from the measurements were also fed to the device
simulation program, which showed reasonable agreement between the simulated
and measured data at lower fluences.
Effects of long term applied bias were also studied using both multi‐electrode charge
readout system (ALiBaVa) and Edge‐TCT. A significant drop in both collected charge
and the leakage current was observed after keeping the detectors under bias for
longer periods of time (> 1000 min). The time evolution of the charge decrease was
found to be fully reproducible under any bias or temperature, influencing long term
annealing induced charge multiplication only. Applying sufficient bias voltage
however (≥ 800 V), results in obtaining a more stable and high enough SNR, providing
optimum detector performance. Both charge collection efficiency and the leakage
current were found to fully recover in late annealing stages after keeping the
detectors at room temperature and no bias for more than 24h.
|
