A High-Granularity Timing Detector for the Phase-II upgrade of the ATLAS calorimeter system: detector concept description and first beam test results

The expected increase of the particle flux at the high luminosity phase of the LHC (HL-LHC) with instantaneous luminosities up to 7.5⋅1034 cm−2s−1 will have a severe impact on the ATLAS detector performance. The pile-up is expected to increase on average to 200 interactions per bunch crossing. The reconstruction performance for electrons, photons as well as jets and transverse missing energy will be severely degraded in the end-cap and forward region. A High Granularity Timing Detector (HGTD) is proposed in front of the liquid Argon end-cap and forward calorimeters for pile-up mitigation. This device should cover the pseudo-rapidity range of 2.4 to about 4.0. Low Gain Avalanche Detectors (LGAD) technology has been chosen as it provides an internal gain good enough to reach large signal over noise ratio needed for excellent time resolution. The requirements and overall specifications of the High Granular Timing Detector at the HL-LHC will be presented as well as the conceptual design of its mechanics and electronics. Beam test results and measurements of irradiated LGAD silicon sensors, such as gain and timing resolution, will be shown.

The expected increase of the particle flux at the high luminosity phase of the LHC (HL-LHC) with instantaneous luminosities up to 7.5·10 34 cm −2 s −1 will have a severe impact on the ATLAS detector performance. The pile-up is expected to increase on average to 200 interactions per bunch crossing. The reconstruction performance for electrons, photons as well as jets and transverse missing energy will be severely degraded in the end-cap and forward region. A High Granularity Timing Detector (HGTD) is proposed in front of the liquid Argon end-cap and forward calorimeters for pile-up mitigation. This device should cover the pseudo-rapidity range of 2.4 to about 4.0. Low Gain Avalanche Detectors (LGAD) technology has been chosen as it provides an internal gain good enough to reach large signal over noise ratio needed for excellent time resolution. The requirements and overall specifications of the High Granular Timing Detector at the HL-LHC will be presented as well as the conceptual design of its mechanics and electronics. Beam test results and measurements of irradiated LGAD silicon sensors, such as gain and timing resolution, will be shown.

K
: Si microstrip and pad detectors; Timing detectors; Electronic detector readout concepts (solid-state); Performance of High Energy Physics Detectors

Design and assembly -module components
The HGTD detector concept is based on individual planar layers of LGAD sensors [2] to be fixed in front of both endcap calorimeter cryostats with active elements between 3435 and 3485 mm in |z|. On both sides of this cooling plate, individual identical modules consisting of LGAD sensors, ASIC and flex cables will be installed. The size of a module is 2×4 cm 2 corresponding to a single LGAD sensor with two ASICs of 2×2 cm 2 each bump-bonded on it. Figure 2 (left) shows three modules with the different components stacked in the z direction. First the ASICs will need to be bump-bonded to the LGAD sensor. This element will then be glued with accurate positioning on the flex cable used to transfer the signals. ASIC signals and HV connections will be performed by wire bonding to the flex cables. On the top and bottom sides of the cooling plate, the position of the modules is in a staggered arrangement This configuration ensures a 20% overlap between modules. The modules disposition and stave orientation are still under optimisation. The modules located at a radius lower than approximately 300 mm are to be replaced after half of the HL-LHC program, due the radiation damage in the sensors and ASICs. An intermediate thin plate will be used to pre-assemble and glue the modules and later screw in the cooling plate to allow a fast dismounting of the inner ring in a long shutdown and re-mounting of new modules. This concept is presented in figure 2 (right) where the green zone is the metallic plate at a radius lower than 300 mm. Yellow and blue squares represent the modules, orange and brown squares represent the flex cables.

Sensors and readout electronics
The time resolution for the HGTD is required to be 30 ps per track over its full lifetime. The sensor pad size is restricted by occupancy, pad capacitance and the fill factor. A unified pad size of 1.3×1.3 mm 2 everywhere with an expected capacitance of 3.4 pF is found to fulfill the requirements. The pads will be arranged in arrays of total area of 2×4 cm 2 with a common backplane bias voltage connection.
LGADs are planar silicon detectors with internal gain. Figure 3 shows one single pad sensor produced at CNM. They are n-on-p silicon detectors containing an extra highly-doped p-layer below the n-p junction to create a high field which causes internal gain as displayed in figure 3. The sensors will be read out by dedicated on-detector front-end electronics ASICs (bumpbonded to the sensors) which should keep the intrinsic excellent time resolution of the LGAD. The contribution to the time resolution from the electronics is given by: where σ jitter depends on the noise and the pulse slope. The requirements of the ASIC, driven by the targeted 30 ps time resolution per MIP after irradiation by combining the multi-hits information, are summarized in table 2. Each pixel readout channel will consist in a preamplifier followed by a discriminator, both defining elements for the overall electronics time performance (figure 4). A first prototype of the ASIC, ALTIROC0, has been designed using the TSMC 130 nm process implementing both the preamplifier, the TOT and a CFD. The prototype chip contains eight channels, four optimized for a C d of 2 pF and four optimized for larger capacitance. Measurements have been performed only on the 2 pF channels using a picosecond generator to provide a voltage test pulse with a rise time smaller than 100 ps. At 10 fC, a jitter of about 25 ps is measured.

HGTD test beam results for non irradiated sensors
During beam tests in 2016 at CERN with 120 GeV pions [3], the gain was measured as a function of the position in the pads by combining the information from the beam telescope with the signal on the LGAD detector (see figure 5). The circular structure in the central part of the single pad sensor has a slightly lower gain than the external part. The efficiency is defined at a given position in the pad as the number of hits that induce a sensor response (with amplitude above threshold) divided by the total number of reconstructed tracks crossing the sensor at that position. The measured 2D distribution for one single-pad sensor is shown in figure 6 (right). The time resolution has been studied in various beam tests. It has been consistently shown that sub-30 ps time resolution can be achieved below the breakdown point before irradiation for the 1.3×1.3 mm 2 LGA devices with about 4 pF capacitance for all gain splits. Measurements based on data from the Autumn 2016 HGTD beam test show a best performance obtained for the ZCD algorithms. Figure 6 (right) presents the time resolution as a function of the gain for single pads and arrays. The best performance is obtained for the sensor with medium doping.

HGTD Preliminary
Test beam Autumn 2016 S1M-1 S1M-2 S1H A3M A2M Figure 6. Efficiency in percent for the single-pad sensor, as a function of the position on the pad (left) and ime resolution using the ZCD method as a function of the gain for single pad sensors and arrays (right).

Sensor performance after irradiation
The gain was found to decrease with irradiation which was attributed to loss of the effective doping concentration in the multiplication layer [4]. This can be observed in figure 7 (left) for CNM devices. The gain steadily decreases with irradiation and after fluences of 6 × 10 14 n eq /cm 2 . There is little difference between standard PIN and LGAD devices. Nevertheless gains of G=8 were measured for devices irradiated to 4 × 10 15 n eq /cm 2 . The gain increases at lower temperatures due to an increase of impact ionization, whereas the breakdown voltage slightly decreases.
The timing performance of CNM LGADs irradiated to different fluences up to 2 × 10 15 n eq /cm 2 is shown in figure 7(right). For HPK LGADs, the time resolution was measured with a 90 Sr setup using the time difference to a calibrated thin LGAD. These results show that the currently available LGAD sensors can be operated safely up to a fluence of 4.5×10 15 n eq /cm 2 , keeping a time resolution of 50 ps, below the required limit of 60 ps per layer, in case of a 4 layers /side layout.

Conclusion
Requirements and design of a High-Granularity Timing Detector (HGTD), to be installed in ATLAS during the shutdown, (end of 2023-mid-2026) have been shown. First prototypes tests in beam tests and in the laboratory have been presented. This is the result of two years of active R&D, especially on sensors and front-end electronics by ∼21 institutes and ∼120 collaborators, started in summer 2015. The HGTD should provide a timing resolution of ∼30 ps for minimum-ionizing particles, covering the pseudorapidity region between 2.4 and 4.0. Intense R&D is ongoing with thinner sensors and different doping materials to increase even further the radiation resistence of the sensors.