Evaluation of slim-edge, multi-guard, and punch-through-protection structures before and after proton irradiation

Planar geometry silicon pixel and strip sensors for the high luminosity upgrade of the LHC (HL-LHC) require a high bias voltage of 1000 V in order to withstand a radiation damage caused by particle fluences of 1 10 1 MeV neq/cm and 1 10 1 MeV neq/cm for pixel and strip detectors, respectively. In order to minimize the inactive edge space that can withstand a bias voltage of 1000 V, edge regions susceptible to microdischarge (MD) should be carefully optimized. We fabricated diodes with various edge distances (slim-edge diodes) and with 1–3 multiple guard rings (multi-guard diodes). AC coupling insulators of strip sensors are vulnerable to sudden heavy charge deposition, such as an accidental beam splash, which may destroy the readout AC capacitors. Thus various types of punchthrough-protection (PTP) structures were implemented in order to find the most effective structure to protect against heavy charge deposition. These samples were irradiated with 70 MeV protons at fluences of 5 10 1 MeV neq/cm–1 10 1 MeV neq/cm. Their performances were evaluated before and after irradiation in terms of an onset voltage of the MD, a turn-on voltage of the PTP, and PTP saturation resistance. & 2012 Published by Elsevier B.V. 1. n-in-p pixel and strip sensor for ATLAS upgrade The Large Hadron Collider (LHC) has been running at CERN in Geneva since 2009 [1]. The LHC is planned to achieve an integrated luminosity of 350 fb . From 2022, the upgraded LHC (HL-LHC) will deliver an instantaneous luminosity 5 times as high as that of the existing LHC. A target number of the integrated luminosity is 3000 fb 1 at the HL-LHC. The ATLAS inner detectors will be replaced with new silicon detectors for the HL-LHC during the 2020 shutdown. In the upgraded ATLAS detector at the HL-LHC, pixel sensors placed 3.3 cm and microstrip sensors placed 30 cm from the beam pipe will be exposed to a radiation fluence of 2 10– 1 10 1 MeV neq/cm respectively. Therefore, we have been developing highly radiation tolerant n-in-p silicon microstrip and pixel sensors for the HLLHC [2]. Elsevier B.V. We have chosen to focus on n-in-p type sensors, since radiation damage does not reverse the bulk type. Lattice defects created by irradiation behave as p-type impurities in bulk silicon. Therefore, radiation damage only increases the p-type carrier concentration [3]. As a result, the n-in-p silicon sensor can always read signals from the strip side. Furthermore, these sensors can be operated under partial depletion conditions, because the depletion region always spreads from the readout strip side. 2. Hamamatsu sensors and proton irradiation tests Various n-in-p test samples were fabricated by Hamamatsu Photonics K.K [4] using float-zone highly resistive silicon, FZ1 or FZ3. FZ3 has a thinner depletion region than FZ1, since FZ3 is processed with deeper backside Pþ implantation. Sample dimensions are 4 4 mm (slim-edge, multiguard) or 10 10 mm (PTP) with a thickness of 150, 200, or 320 mm. PTP samples also have a P-stop (Pþ is implanted between strips) or P-spray (Pþ is sprayed on surface) strip isolation structure for preventing from Table 1 Variety of slim-edge and multi-guard samples. 1st test, slim-edge (p-/n-edge) samples 200 mm 320 mm N-bulk FZ1 FZ1 1st test, multi-guard (p-/n-edge) samples 200 mm 320 mm P-bulk – FZ1 N-bulk FZ1 FZ1 2nd test, slim-edge (p-edge) samples 150 mm 320 mm P-bulk FZ1, FZ3 FZ1 P-bulk, P-spray FZ1 FZ1 2nd test, multi-guard (p-/n-edge) samples 150 mm 320 mm P-bulk FZ1, FZ3 FZ1 P-bulk, P-spray FZ1 FZ1 Fig. 1. Comparison of the Full Depletion Voltage for present and previous measurements. Fig. 2. Temperature dependence of leakage current in BZ4D-5 irradiated at 1 10; 1 MeV neq/cm. Fig. 3. Fluence dependence of leakage current in PTP samples. S. Mitsui et al. / Nuclear Instruments and Methods in Physics Research A 699 (2013) 36–40 37 conduction between strips due to an electron accumulation layer at the Si–SiO2 interface induced by positive charges in SiO2 insulation layer. We performed two proton irradiation tests with the 70 MeV beam at the Cyclotron and Radioisotope Center (CYRIC) at Tohoku University [5] (Table 1). During the first test, we irradiated the PTP, slim-edge, and multi-guard ring samples of 200 and 320 mm in thickness at 5.7 10, 1.1 10, 1.2 10, 1.2 10, and 1.1 10 1 MeV neq/cm. All samples were annealed at 60 1C for 80 min after irradiation. During the second test, we irradiated other slim-edge and multi-guard ring samples of 150 and 320 mm in thickness at 1.1 10, 1.2 10, 5.7 10, 1.2 10 1 MeV neq/cm . These samples were then annealed at 60 1C for 65 min equivalent after irradiation. The fluences of 1 MeV neutrons is estimated from multiplying the fluences of 70 MeV protons by 1/0.7 by the NIEL hypothesis. We evaluated the full depletion voltage (FDV), energy gap (Eg) and damage constant for the PTP samples in order to estimate the consistency of the proton irradiation. FDV is estimated from body capacitance of the sensor d1⁄4 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2eV e NAþND NAND s , Cbulk 1⁄4 e S d 1⁄4 S ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ee