The general operational principle of silicon sensors is based on use of the p-n junc- tion [24]; connection of the two doped silicon regions with different concentration of free electrons and holes. The n-doped region represents an area (usually doped with Arsenic, As) with free electrons and the p-doped region represents an area (usually doped with Boron, B) with free holes, thus a bipolar diode is created.
Simple schematic is presented in Figure 1.10.
p n
Si
As Si
Si
Si
-q
Si
B Si
Si
Si
+q
Figure 1.10: p-n junction.
Since the sides of the junction contain excess of electrons or holes a diffusion current flows in both direction; electrons and holes are diffusing leaving behind a region of fixed ions. This region is called the Depletion Region. To extend the depletion region of the sensor an external electrical potential is applied to the junction. See schematic in Figure 1.11.
n p + + + + + + +
+ + + + + + +
- - - - - - -
- - - - - - - - - - - -
- - - - - - - - - -
- - - - - - - - - -
- - - - - - - - - -
- - - - -
+ + + + + + + + + + + + + + +
+ + + + + + + + + +
+ + + + + + + + + +
+ + + + +
n p + + + + + + +
+ + + + + + +
- - - - - - -
- - - - - - - - - - -
- - - - - - -
- - - - - - -
- - - - - - -
- - -
+ + + + + + + + + + +
+ + + + + +
+ + + + + + +
+ + + +
+ + + + + + + + + + + + +
- - - - - - -
- - - - - - - Depletion Region
+ -
Figure 1.11: Depletion region.
The voltage necessary to increase the width of the depleted region (sensitive region of the sensor) is called the Bias Voltage. The depleted region of the sensor is used for particle detection; electrical charges created by the passage of an ionizing particle through the silicon are drifting to the respective electrodes (electrons into n direction and holes into p direction) forced by the electrical field formed in this area and are detected by the readout electronics connected to the silicon sensor. The advantage of the silicon technology is that energy necessary for the electron-hole pair production in silicon is much less (≈3.6 eV) than in ionizing gases (≈30 eV) or diamond (≈13 eV) therefore giving bigger readout signal directly proportional to the released energy and allowing lower energy particle detection. From the other hand diamonds are more resistive to the radiation damage but their production is much more difficult and expensive.
High luminosity in the LHC (integrated radiation dose in the ATLAS ID) has significant effect on the performance of the Pixel and SCT silicon sensors:
displacement of the atoms in the silicon lattice changes effective doping con- centration therefore increases the depletion (bias) voltage for the sensors; The
leakage current of the silicon sensor as well increases with the radiation causing an overall increase in the dissipated power from the sensor affecting the subse- quent change in the silicon doping concentration and the increase in the required operation voltage (depletion voltage); The Pixel [25] and SCT [26] silicon sensors are designed to cope with these requirements. In the ATLAS inner detector they are operated in the temperature range -5◦C to -10◦C (Section 2.2) to minimise the change in the effective doping concentration and lower the leakage current and can guarantee adequate signal performance over the entire inner detector operation period with the given LHC luminosity.
The pixel silicon sensor represents the 250àm thick detector; the array (pixels) of the bipolar diodes implanted into the n-type bulk. High positive (p+) and negative (n+) doped regions are implanted on both sides of the wafer. This allows good charge collection efficiency even after the p-n type inversion caused by the radiation damage [25] . On each sensor 47232 pixel implants with the nominal size of 400::24ấ17::m2 are arranged in 144 columns and 328 rows and in each column eight pairs of pixel implant are ganged resulting in 46080 pixel read-out channels, or 320 independent read-out rows, allowing connection of the sensor tile to the 16 read-out electronic front-end chips. For the quality assurance tests all read-out channels were connected to the common bias grid but in normal operation biasing for each individual pixel is provided by the bump-bond technique [25] through the openings in the passivation layer of the sensor. The nominal bias voltage required for the sensor operation is≈150 V and the maximum predicted bias voltage, after the maximum expected irradiation with respect of the integrated luminosity and the expected operating temperature profile is ≈600 V. The Pixel sensor design guarantees each pixel isolation and a special p-spray isolation technology [25]
was used to allow small feature size and good performance of the sensor after the irradiation.
The SCT sensors are 285±15àm thick silicon detectors with the standard single sided p-n technology, the p-strips on the high resistive n-bulk, with the
AC-coupled readout strips. In total there are 15912 silicon sensors used in the SCT sub-detector. For the SCT barrel sensors 768 readout strips are positioned with 80àm pitch but for the EC sensors are not at a constant pitch since the EC sensors have a wedge-shaped geometry in contrast to the barrel rectangular sensors [26]. The complicated sensor geometry was adopted because of their layout on the EC discs. The different type (geometry) EC sensors are grouped in five and named W12, W21, W22, W31 and W32 to form the sub-detector modules [26]. Each sensor of both the SCT barrel and EC modules are read out by six 128-channel ABCD3TA ASICs [27]. The nominal bias voltage required for the SCT sensor operation, as it was proposed in the ATLAS TDR [6], is
≈150 V and the predicted bias voltage range after the irradiation at the end of the ATLAS inner detector operation period is between 350 V and 450 V depending on the position of the sensor, integrated luminosity and the expected operational temperature profile. These values were revised later in this work in Chapter 3.