Rousseau El Emilio

Nuclear Physics B (Proc. Suppl.) 215 (2011) 151–153 www.elsevier.com/locate/npbps

Silicon Detectors for the sLHC J. Metcalfea , on behalf of the RD50 Collaboration. a University of New Mexico MSC074220 Department of Physics and Astronomy 800 Yale Blvd. NE Albuquerque, NM, 87131, USA

The luminosity upgrade of the Large Hadron Collider (LHC) at CERN to the Super LHC (sLHC) will increase the radiation dose at the experiments by roughly an order of magnitude. The elevated radiation levels require the LHC experiments to upgrade their tracking systems with extremely radiation hard detectors. Recent results on defect characterization of silicon materials and radiation hard technologies developed by the RD50 Collaboration for sLHC use are reported. Studies on n- and p-type silicon with Float Zone (FZ) and Czochralski technologies as well as 3D silicon detector designs are presented. Properties such as charge collection efficiency, electron signal, effective doping concentration (Neff ) and full depletion voltage (Vfd ) after radiation exposure are compared to assess the performance of the different technologies.

1. Introduction The RD50 Collaboration is a CERN based organization aimed at the development of radiation hard semiconductor devices for very high luminosity colliders. The current silicon sensor technologies (i.e. planar n-type Float Zone in ATLAS at the LHC) may not operate effectively at the Super LHC where fluences at the innermost layers are expected to reach around 1.6 x 1016 1 MeV neq /cm2 . RD50 has made major advancements to meet the challenges of the Super LHC environment. These include understanding the underlying cause of radiation damage to silicon materials through microscopic defect behavior, characterizing macroscopic defects due to radiation damage and investigating new silicon detector technologies such as p-type, Czochralski and 3D silicon detector materials. A more complete description of all research by RD50 members can be found at http://rd50.web.cern.ch/rd50/. 2. Radiation Damage Mechanisms 2.1. Microscopic Defects A comprehensive understanding of the causes and effects of microscopic defects produced by different types of radiation damage promotes a 0920-5632/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.nuclphysbps.2011.03.162

more efficient method to engineer new silicon materials for radiation hard applications. The two main types of microscopic defects are surface damage and bulk (crystal) damage. Surface damage due to Ionizing Energy Loss (IEL) leads to an accumulation of positive charge in the oxide (SiO2 ) and silicon-oxide (Si/SiO2 ) interface, which affects the inter-strip capacitance and the breakdown behavior of the sensor. Bulk damage is caused by displacement damage resulting in NON-Ionizing Energy Loss (NIEL), which leads to the build-up of crystal defects. The WODEAN (Workshop On DEfect ANalysis) project was established by RD50 member Gunnar Lindstroem and includes 10 RD50 member institutes to identify the individual defects responsible for trapping, leakage current and changes in Neff . Approximately 240 identical samples were irradiated by protons and neutrons and measured using numerous techniques: Capacitance/Current Deep Level and Photo Induced Transient Spectroscopy, Fourier Transform Infrared Spectroscopy, Thermally Stimulated Currents, Recombination Lifetime, Photo Conductivity, Electron Paramagnetic Resonance, Transient Charge Technique and CV/IV. Many point defects and cluster related centers have

152

J. Metcalfe / Nuclear Physics B (Proc. Suppl.) 215 (2011) 151–153

been identified as resulting from different types of radiation damage and linked to changes in macroscopic properties such as Neff , leakage current and annealing behavior. Several important results are published in [1–4]. For example, epitaxial silicon detectors were exposed to proton and neutron radiation. The defects resulting from the proton irradiation were shown to add more positive space charge to the silicon while the neutron irradiation introduced defects that increased the negative space charge. The changes in the space charge and hence the effective doping concentration, Neff , are consistent with direct measurements of Neff –a macroscopic property. 2.2. Macroscopic Defects Radiation damage affects the build up of positive or negative space charge as described in the previous section. The distribution of space charge in the detector determines the electric field profile across the detector. This in turn affects the depletion voltage, signal and charge collection efficiency. The electric field profile can be measured using the Transient Current Technique (TCT). In one study, the TCT was applied to Magnetic Czochralski (MCz) silicon sensors irradiated with neutrons followed by photons. In this case it is known that neutrons produce negative space charge while photons induce positive space charge in MCz. It was shown that the introduction of opposite signs of space charge balance each other out (although some indication of interaction between the defects was seen at higher fluences) [5]. The accumulation of space charge also predicts the annealing behavior. For silicon samples with negative space charge after irradiation, there is an initial period of beneficial annealing (corresponding to a decrease in Vfd ) while positive space charge is introduced into the detector followed by reverse annealing. The opposite occurs to sensors that have a positive Neff after irradiation–reverse followed by beneficial annealing [6]. 3. New Silicon Detector Technologies 3.1. p-type Silicon Members of the RD50 collaboration are studying p-type (n-in-p and n-in-n) silicon materials.

In p-type Si sensors electrons are collected by the electrodes on the n-side instead of holes from electrodes on the p-side as in n-type (p-in-n) Si sensors. The faster electron mobility yields lower trapping probability and larger collected signal. p-type sensors have the added benefits that after irradiation there is no type-inversion and that the single-sided processing reduces production cost. Several studies have been made to compare signal and charge collection efficiency of various pand n-type silicon materials. These studies show that p-type sensors indeed have a higher signal than n-type at the same bias voltage [7–10]. The signal-to-noise ratio of p-type sensors is fairly stable during annealing [11]. This would provide the added benefit that stable operation of the detector could be maintained during maintenance periods without additional cooling. 3.2. Czochralski Silicon Czochralski silicon materials are also under investigation. Silicon sensors made using the Czochraski (Cz) and similarly the Magnetic Czochraslki have a higher oxygen content making the formation of shallow thermal donors possible after irradiation, which reduces the bias voltage required for full depletion voltage. The Cz and MCz are both shown to require less bias voltage for full depletion after irradiation than standard oxygenated Float Zone (FZ) silicon [6,12]. When MCz is exposed to mixed irradiations (first protons or pions followed by neutrons) the space charge introduced by the radiation compensates in MCz silicon materials allowing for the recovery of the Vfd whereas in FZ the effects are additive and Vfd continues to degrade regardless of the type of radiation particle [13]. 3.3. 3D Silicon Silicon detectors with columnar electrodes through the bulk of the detector instead of on the surface as in the more traditional planar design are categorized as 3D sensors. The design of 3D has the advantage over planar that the shorter distance between electrode columns means the electrons have a shorter distance to travel to the electrodes for charge collection. This reduces trapping and lowers the bias voltage re-

J. Metcalfe / Nuclear Physics B (Proc. Suppl.) 215 (2011) 151–153

quired for full depletion. Members of RD50 are investigating a double-sided 3D design with alternating rows of n- and p-type columns where the n-type columns are drilled from the top surface of the sensor into the bulk and p-type are bored from the bottom. This design is manufactured at two locations–CNM (Centro Nacional de Microelectronica) in Barcelona and FBK (Fondazione Bruno Kessler) in Trento. Recent results from irradiated 3D sensors show signal collection to be very promising with more than 20,000 electrons after irradiation to 2 x 1015 neq /cm2 for bias voltages less than 200 V [14–16]. Another 3D design is also underway by RD50 collaborators at Brookhaven National Laboratories called the Independent Coaxial Detector Array (ICDA). Again, this design uses a 3D trench electrode model, but with an asymmetrical electrode configuration, which produces a near homogeneous electric field without any theta dependence around the electrode [17,18]. 4. Summary The knowledge gained from the study of microscopic and macroscopic defects in silicon after radiation exposure has narrowed the field of focus for new types of silicon detectors for application in the sLHC and other high luminosity colliders. Both p-type and Czochralski planar silicon technologies as well as 3D silicon are currently under investigation by the RD50 collaboration. A comparison of the signal collected by each technology is shown in Figure 1 [13,19–24]. While Figure 1 provides valuable insight to compare the performance of each technology, there are many other factors to consider in selecting a suitable technology such as signal-to-noise, infrastructure and cooling requirements–all of which must be weighed according to the needs of a given application. It is apparent that these silicon detector technologies demonstrate readiness to meet the performance requirements of the sLHC environment. REFERENCES 1. I. Pintilie, et al., APL 92, 2008.

153

Figure 1. The signal of different silicon technologies (and diamond) are compared as a function of fluence. Not included: CNM 3D sensors show more than 20,000 electrons after irradiation to 2 x 1015 neq /cm2 .

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

J. Stahl, et al., Physica B 340-342 (2003) 705. A. Junkes, et al., NIMA 612, (2010) 525. I. Pintilie, et al. NIMA 611 (2009) 52. Z. Li, et al. IEEE NSS conf. rec. N15-6 (2008). J. Metcalfe, RD50 Workshop June 2009. C. Tosi, et al., NIMA, 579 (2007) 766. J. Bohm, et al., NIMA, in press, (2010). A. Macchiolo, et al., NIMA 591(2008) 229. L. Andricek, et al., NIMA in press, (2010). G. Casse, et al., NIMA 568(2006) 46. M. Bruzzi, MPGD 2009. G. Kramberger, et al., NIMA 609 (2009) 142. M. Koehler, RD50 Workshop June 2010. G. Pellegrini, et al., NIMA, 592 (2008) 38. A. Zoboli, et al., IEEE TNS, 55 (2008) 2275. Z. Li, et al., NIMA 583 (2007) 139. Z. Li, RD50 Workshop June 2010. G. Casse, et al., IEEE TNS 55 (2008) 1695. M. Petasecca, et al., NIMA 563 (2006) 192. W. Adam, et al., NIMA 565 (2006) 278. C. Fleta, et al., NIMA 579 (2007) 642. G. Lindstroem, et al., NIMA 568 (2006) 66. D. Creanza, et al., NIMA 579 (2007) 608.