Speaker
Description
For medium and highly irradiated silicon devices such as
strip detectors or diodes it has been observed that the commonly used CV
method for deriving depletion voltage as well as doping level is not as easily
applicable as for unirradiated devices. The reason for the arising difficulties is
that defects created in silicon can capture and release charge
carriers and therefore show a time dependency which affects the CV
measurement. To be able to measure highly irradiated sensors a setup has
been designed and assembled which enables a CV frequency variation down to
20Hz and temperature control to get down to -20 to -40°C to
investigate the influence of the defects superimposing the geometrical
behaviour of the devices. With the measured data an impedance
spectroscopy has been performed to differentiate between a
capacitive (the ideal case) and a resistive behaviour which allows
a determination of an effective time constant for the defects present.
Lowering the temperature offers the possibility to increase
this time constant according to Boltzmann's law, and by this an
effective energy depth of trap levels can be calculated.
Comparison with unirradiated devices indicates the effects introduced by
radiation damage and comparison of diodes with strip detectors shows the
role of the segmented front side.
As expected, strip detectors show at high frequencies a deviation from a
pure capacitive behaviour which can be explained by the RC element
caused by bias resistors. For unirradiated devices no
temperature dependence is observed as it would be expected for a constant resistance
whereas for irradiated sensors we observe a strong temperature dependece
which can only be explained by either a temperature dependent resistance
or some interaction with defects in the silicon.
Furthermore, it is shown which frequencies of the CV measurements are an optimal choice for different analysis goals like doping or effective energy
depth determination.
