Recent interest in pile-up mitigation through fast timing at the HL-LHC has focused attention on technologies that now achieve minimum ionising particle (MIP) time resolution of 30 picoseconds or less. The constraints of technical maturity and radiation tolerance narrowed the options in this rapidly developing field for the ATLAS and CMS upgrades to low gain avalanche detectors and silicon photomultipliers. In a variety of applications where occupancies and doses are lower, devices with pixel elements of order 1 cm 2 , nevertheless achieving 30 ps, would be attractive.In this paper, deep diffused Avalanche Photo Diodes (APDs) are examined as candidate timing detectors for HL-LHC applications.Devices with an active area of 8 × 8 mm 2 are characterised using a pulsed infrared laser and, in some cases, high energy particle beams. The timing performance as well as the uniformity of response are examined.The effects of radiation damage on current, signal amplitude, noise, and timing of the APDs are evaluated using detectors with an active area of 2 × 2 mm 2 . These detectors were irradiated with neutrons up to a a 1-MeV neutrons fluence Φ eq = 10 15 cm −2 . Their timing performance was characterised using a pulsed infrared laser.While a time resolution of 27 ± 1 ps was obtained in a beam test using an 8 × 8 mm 2 sensor, the present study only demonstrates that gain loss can be compensated by increased detector bias up to fluences of Φ eq = 6 · 10 13 cm −2 .So it possibly falls short of the Φ eq = 10 14 cm −2 requirement for the CMS barrel over the lifetime of the HL-LHC.
IntroductionThe high luminosity upgrade of the CERN Large Hadron Collider (HL-LHC) foreseen to start in 2026 will provide an instantaneous luminosity of up to 5 · 10 34 cm −2 s −1 with a bunch spacing of 25 ns, and an average pile-up of up to 200 collisions per bunch crossing [1]. This value of pile-up presents a challenge for the experiments, as currently ATLAS and CMS have reached a typical number of concurrent interactions within the same bunch crossing (pile-up) of approximately 35 [2, 3]. At present, the effects of pile-up on physics analyses are mitigated by resolving the primary vertices within one bunch crossing along the beam axis. The physics objects are then associated to the correspondingvertices by using the tracking information. This strategy was used also at Tevatron, where the pile-up was ≈ 6. The pile-up at HL-LHC will pose a challenge to this method, as a significant fraction of the primary vertices will have a distance smaller than the resolution of the vertex detectors, making an association to the reconstructed physics objects impossible.A different method to associate the reconstructed objects to separate vertices relies on the measurement of the time of arrival of the particles at the detectors.A sufficiently accurate time measurement effectively reduces the vertex density, improving the event reconstruction capability of the experiments. Since the RMS spread of the primary vertices at HL-LHC is foreseen to be ≈ 170 ps within on...