ATLAS Upgrade SSD Project:

Transcription

1 ATLAS Upgrade SSD: Specifications of the ATLAS Upgrade SSD ATLAS Project Document No: Institute Document No. Created: 30/10/2006 Page: 1 of 7 DRAFT 1.0 Modified: ATLAS Upgrade SSD Project: Specifications of the ATLAS Upgrade SSD ATLAS Upgrade Document no.: Contact persons: Prepared by: Phil Allport, Liverpool U. (allport@hep.ph.liv.ac.uk), Manual Lozano, CNM Barcelona (manuel.lozano@cnm.es), Hartmut Sadrozinski, UC Santa Cruz (hartmut@scipp.ucsc.edu), Yoshinobu Unno, KEK (unno@post.kek.jp) Checked by: Approved by: Distribution List

2 ATLAS Project Document No: Page: 2 of 13 History October 2006 November 26, 2006 Order 0 draft oder 1 draft: PPA, YU, HFWS commented,

3 ATLAS Project Document No: Page: 3 of 13 1 PURPOSE These specifications define the environment, the geometrical layout and the expected performance of silicon strip detectors (SSD) being developed for the upgrade of the ATLAS Inner Detector. They will allow comparison of the performance of prototype SSD manufactured by different vendors and on different wafer types. 2 SCOPE After operation of about 5-10 years at full luminosity at the LHC, the ATLAS Inner detector (ID) will have to be replaced because of radiation damage. In addition, there is a plan to upgrade the LHC with a luminosity increase. The requirements for the SSD become much more stringent when compared to the LHC, requiring a survival fluence level of about neq/cm 2, a factor of 10 increase with respect to the present LHC [1]. Both short and long strips are covered. These specifications are developed in the context of the Inner Detector (ID) layout and need to be updated if the ID layout changes or technical solutions change. Another connection exists to the module R&D program, to the pixel development and the front-end electronics R&D program. 3 DEFINITIONS 3.1 Acronyms 60 CO Cobalt gamma source ASIC Application Specific Integrated Circuit DUT Device under Test ID Inner Detector SCT (ATLAS) SemiConductor Tracker SSD Silicon Strip Detector TID Total Ionizing Dose TKR Tracker 3.2 Definitions and Abbreviations cm centimeter krad 1000 Rad MeV Million electron Volt MRad 10 6 Rad N neutron P proton Rad unit of TID s, sec second y year

4 ATLAS Project Document No: Page: 4 of 13 4 APPLICABLE DOCUMENTS 4.1 ATLAS & LHC Documents SLHC Physics: F. Gianotti, M.L. Mangano, T. Virdee, et al. CERN-TH/ (April 1, 2002) SLHC Machine: O. Bruhning et al., LHC Project Report 626SLHC 4.2 References [1] Ian Dawson, [2] A. Abdesselam et al, The Barrel Modules of the ATLAS SemiConductor Tracker, accepted for publication by Nucl. Instr. and Methods A. [3] A. Abdesselam et al, The ATLAS SCT Endcap Module, to be submitted to Nucl. Instr. And Methods A. [4] M. Bruzzi, Radiation Damage in Silicon Detectors for High-Energy Physics Experiments, 6th International Hiroshima Symposium on the Development and Application of Semiconductor Tracking Detectors, September 2006, Carmel, CA, USA, [5] M. Moll et al., Leakage current of hadron irradiated silicon detectors material dependence, Nucl Instr Meth A 426 (1999) [6] H. F.-W. Sadrozinski, A. Seiden, M. Bruzzi, Operation of Short-Strip Silicon Detectors based on p-type Wafers in the ATLAS Upgrade ID, SCIPP 05/09 [7] Y. Unno, Sensor Options [8] S. Terada et al. Proton Irradiation on p-bulk Silicon Strip Detectors using 12GeV PS at KEK, Nucl. Instr. and Methods A383 (1996) p [9] M. Hanlon, Ph.D Thesis (Liverpool) The Development of P-type Silicon Detectors for the High Radiation Regions of the LHC (1998) [10] [CERN-RD50 collaboration, Radiation hard semiconductor devices for very high luminosity colliders, [11] Y. Unno et al., "Evaluation of p-stop structures in the n-side of n-on-n silicon strip detectors", IEEE Trans. Nucl. Sci. Vol. 45, pp , 1998, [12] Y. Unno et al., "Novel p-stop structure in n-side of silicon microstrip detector", Nucl. Instr. Meth. A541 (2005), [13] ATLAS Inner Detector Technical Design Report, CERN/LHCC/97-16 and CERN/LHCC/97-17, [14] CMS TDR, 1998, CERN/LHCC 98-6 CMS TDR. [15] LHCb-VELO TDR, CERN/LHCC 31 May [16] E. Tuovinen et al., N and P-type Cz-Si Detectors irradiated with high and low energy protons, 4th RD50 - Workshop on Radiation hard semiconductor devices for very high luminosity colliders, CERN 5 May, [17] G. Kramberger, Annealing of effective trapping times in irradiated silicon detectors, 6th International Hiroshima Symposium on the Development and Application of Semiconductor Tracking Detectors, September 2006, Carmel, CA, USA, [18] Y. Unno, "Silicon sensor development for the ATLAS upgrade for SLHC", Nucl. Instr. Meth. A569(2006),41-74 [19] G. Casse, P.P. Allport, S. Marti y Garcia, M Lozano, P.R. Turner, "Performances of miniature micro-strip detectors made on oxygen enriched p-type substrates after very high proton irradiation" Nuclear Instruments and Methods in Physics Research A, vol. 535/1-2, (2004) [20] H.F.-W. Sadrozinski et al., Total Dose Dependence of Oxide Charge, Interstrip Capacitance and Breakdown Behavior of slhc Prototype Silicon Strip Detectors and Test Structures of the SMART Collaboration, 6th International Hiroshima Symposium on the Development and Application of Semiconductor Tracking Detectors, September 2006, Carmel, CA, USA, [21] 3rd RD48 STATUS REPORT, CERN LHCC

5 ATLAS Project Document No: Page: 5 of 13 5 Specifications of the Upgrade SSD 5.1 Environment for the Upgrade SSD Layout of the Upgrade Inner Detector (ID) The all-silicon tracker is expected to involve at least three different technologies, pixels, short micro-strips and long micro-strips as shown in the Fig. 1. Fig.1 Straw man layout of the upgrade ID Possible radial dimensions and sensor geometries are shown in Table I below, and more discussion on alternative layouts is provided in Sec A further consideration is the sensor requirements for the discs, where many different sensor types seems inevitable if the strips are to measure the phi coordinate. The technology for barrel layer 3, where the estimated hit occupancies exceed 1%, is also under discussion, with possibilities including a standard pixel layer or a layer with large pixels or very short strips. For the short-strip layer, a maximum instantaneous luminosity of cm -2 s -1 implies for an average hit rate of < 1% and for similar sagita resolution to the SCT, either single-sided strip dimensions of roughly 3 cm by 50 µm or double-sided small angle stereo mounted back-to-back of 80 µm pitch (as currently in the SCT [2,3]) A finer segmented layer may also be required for the 27cm layer and this may be beyond the radii where the current hybrid pixel technology can be cost effective. In the case where the short-strip layers primarily serve for momentum measurements, additional large pixel or very short strip layers for robust pattern recognition could be required in the transition region between the pixel vertex detector and the micro-strip tracking. Table Ia: Dimensions in the Barrel of the Upgrade ID

6 ATLAS Project Document No: Page: 6 of 13 Table Ib: Dimensions in the Endcaps of the Upgrade ID (inner radii are for η = 2.5 from 100 mm away from the center) Radiation Levels For estimating the radiation requirements we assume 3000fb -1 and a safety factor of two. Using the parametrization of Ref [1], this leads to the fluences shown in Fig. 2, which also shows the location of the layers. To account for the z-dependence, the fluences for pixels and short strips will be multiplied by 1.115, while for the outer radii, the fluences for z=150 cm are taken, which are about 15% higher than in Fig. 2 (see Table III) [1]. The requirement of tolerating for the short strips is neq/cm 2 (including the 2x safety factor). The higher radius long micro-strip layers from 70cm outwards require tolerance to doses of about neq/ cm 2 i.e. about double the current SCT. Pixel dose estimates range from ~10 16 neq/ cm 2 (depending on radius) for the b-layer to doses of a few neq/ cm 2 at higher radii. The split of doses between neutrons and pions and protons is important, since the use of oxygen enhanced substrates (eg DOFZ or MCz) is much more beneficial for protons. There is little data on radiation damage due to pions. At high radius, the dose is, however, dominated by neutrons. Fluence neq/cm 2 1.E+17 1.E+16 1.E+15 1.E+14 satlas Fluences for 3000fb-1 All: RTF Formula All: 2x Margin n (5cm poly) pion proton 1.E+13 1.E Thermal Environment Radius R [cm] Fig. 2 Fluence in 1 MeV neutron equivalent as a function of radius [1]. The thermal management has to deal with the control of the radiation induced detector damage [4]. Besides the leakage current, which without proper cooling could lead to excessive noise and thermal run-away, the temperature must be sufficiently low to prevent the increase in depletion voltage known as anti-annealing. The leakage current I increases lineraly with fluence F [5]: 120

7 ATLAS Project Document No: Page: 7 of 13 I = α Φ Volume with a leakage current damage constant for 1 MeV neutrons α n = A/cm at 20 C after canonical annealing. If T 0 = 293K, the current at T is parameterized as: 2 T E 1 1 ( ) ( 0) exp( ) 0 2 b I T = I T T K T0 T with E b =1.12 ev silicon band gap. This leads to the temperature dependence of the leakage current current damage constant shown in Table II. Table II: Relative leakage current constants α n for different temperatures T Temperature T [ o C] α(t)/ α(20) For end of life, the expected leakage current per detector element (strip or pixel) of the stray man layout (Table I) is shown in Table III, together wit the expected fluence. This is important for noise calculations. The expected power density for 3 temperatures (0, -10, -20 o C) is shown in Table IV, assuming a maximum bias voltage of 600V, as justified in Section 5.3 and Ref. [6]. These numbers can be used for assesment of the risk of thermal run-away. Table III: Fluence and Leakage Current/channel vs. Radius for different Temperatures T (Integrated Luminosity: 3000fb -1 ) Leakage current I [A] per strip/pixel for temperature T [ o C] R [cm] F [neq/cm 2 ] E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E (12cm) 2.0E E E E E E E E (12cm) 1.5E E E E E E E E-07 Table IV: Fluence Power Density vs. Radius for different Temperatures T (for 3000fb -1 and assuming 600V bias voltage) Power Density [W/mm 2 T [ o C] R [cm] E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E-05

8 ATLAS Project Document No: Page: 8 of Geometry 5.2 Layout of the Upgrade SSD The dimensions of the sensors are listed in Table I. The largest area that can be safely made available from a 15 cm wafer is cm by cm [7] corresponding to 4 rows of 3cm strips (plus bias and guard regions) giving, at 80 µm pitch, 3072 read-out channels (with 50 µm pitch, this would need 10 rows of 128 channel read-out ASICs or 5120 channels). Each wafer would then require 24 (40) 128-channel front-end ASICs. The choice could be between back-to-back layers at 80 µm or 50 µm pitch single-sided (but with one or more high radius pixel or very short strip layers at lower radii). However, the ASIC designers will have difficulty bringing all their connections out the back as required for the 50 µm case. Therefore prototyping with early ASICs or existing chips will focus on the 80 µm case. For the outer layers, the options of strip lenth are 9cm vs. 12cm Wafers Pixel layers using the current n-in-n technology are expected to prove adequate for all but the b-layer and the starting point here, for the sensors, should be the current designs. These could be imported into an n-in-p [8,9] technology, using the same masks if desired. There is the issue of needing guard rings on the back-side to prevent the bias voltage to appear on the implant side and short out the ASICs. At large radius, the following options exist: sensors based on current p-in-n technology vs. newly developed n-inp, and 12cm vs. 9cm length strips. The additional radiation tolerance requirements need to be verified. The radiation hardness required at the SLHC, for all but the 5 cm b-layer, should be achievable by using technologies developed by the RD50 collaboration [10] and the effort lead by Japanese groups [11], [12]. For the short-strip region float-zone (FZ) p-type (with n-side read-out) would be a conservative route to charge collection efficiencies of 70% of the unirradiated values in a single-sided technology. Other options include n-type readout strip in n-type silicon as used in the ATLAS pixel system but this would require a potentially more expensive double-sided lithographic processing [13,14,15]. A more exotic substrate option would be high resistivity Magnetic Czochralski (MCz) which is believed to reduce changes of substrate doping (and the associated rise in required operating voltage) with dose [16]. This material looked promising that it could even be used with the more conventional p-strips in n-type silicon, because earlier studies suggested that it did not invert from n to p type. Recent results indicate inversion for neutron and high fluence proton irradiations and need to be confirmed [4], This could be particularly attractive for the long microstrip layers. However, for the short-strip layers, since such a configuration involves collecting holes, the charge collection efficiency would be expected to be more susceptible to trapping than n-side read-out collecting electrons [17]. Since radiation will increase the depletion voltage V FD monotonically, the lowest practical initial depletion voltage is required. A specification of between 50V < V FD < 100V for 300 µm thick detectors is achievable with as processed wafer resistivity ρ of 10kΩ-cm < ρ < 20k Ω-cm, corresponding to a doping density N eff of 7*10 11 cm -3 < N eff < 1.4*10 12 cm Design Details Biasing As in the present SCT, biasing with AC-coupled sensors would be achieved with polysilicon resistors with resistance of about 1MΩ. For DC-coupled sensors, a punch-through biasing network could be envisioned, to facilitate testing Coupling caps As in the present SCT, coupling capacitors on AC-coupled sensors would be using both silicon oxide and nitride Strip isolation

9 ATLAS Project Document No: Page: 9 of 13 A combined combined p-spray and p-implant isolation for p-type sensors has been succesfully prototyped with HPK [18] and is being tested in the RD50 submission, promising both good performance pre-rad (low p-spray) and post-rad (implants) Guard rings/ Edge treatment The guard ring(s) will be on the strip side only. The details are left to the manufacturer. 5.3 Expected Performance of the Upgrade SSD The results of the charge collection efficiency measurements performed with LHC-speed analogue electronics show the ability of these devices to survive the anticipated SLHC radiation doses. Figure 3 shows the deterioration of the charge collection efficiency as a function of operating voltage at different 24GeV/c proton fluences [19]. Even at the very high fluences, corresponding to roughly 10 cm radius at SLHC (equivalent to p cm -2 ) the collected charge is as high as 7000 electrons. This could be sufficient for the pixel detector at SLHC provided the noise can be limited to at least a factor of ten smaller than the signal. N-in-n sensors, which also work by collecting electrons, are expected to give equivalent performance. For the strip detectors, a bias voltage between 600V and 700V is sufficient for the survival neutron equivalent fluence of neq cm -2 coresponding to a proton fluence of p cm -2. Figure 3: Collected Charge as a Function of Bias Voltage at Different Proton Fluences for p-type Substrate Miniature Detectors (1 cm 2 ) [19]. The noise will depend on the ASIC technology used. Based on experiences with the SCT readout, the noise can be parametrized as [6] σ 2 Noise = (A + B C) 2 + (2 I τ s )/q where C is the total capacitance, and the last term is due to the leakage current I (q = C). For strips, we will adopt ATLAS SCT like numbers A = 500 e -, and B = 50 e - /pf, and the capacitance is parameterized as C=L*1.5pF, where L is the strip length (3.5cm for SS, 9cm for LS), and the capacitance per cm has been measured on a variety of p-type detectors [20]. An ATLAS SCT shaping time of τ s = 20 ns is assumed both for pixels and strips. Table III show the noise per strip for the different detector radii. For the pixel detectors, only the noise due to leakage current is shown. It has to be added in quadrature to the pre-rad noise contribution. For strips at radius 27 cm, the noise is decreased by about 22% (3000fb -1 ) and 38% (6000fb -1 ) by decreasing the operating temperatures from -10 o C to -30 o C, and the signal-to-noise ratio increases from about 18.6 to 22.7 (3000fb -1 ) and 13.2 to 18.2 (6000fb -1 ). In the outer layer at 70cm radius, decreasing the temperature from -10 o C to -30 o C, the signal-to-noise improves from 16.7 to 18.4 for 9cm log strips and from 14.1 to 15.4 for 12cm long strips, for 3000fb -1, and 13.8 to 16.4 for 9cm log strips and from 11.7 to 13.8 for 12cm long strips, for 6000fb -1.

10 ATLAS Project Document No: Page: 10 of 13 Table III: Strip Noise pre-rad and after 3000fb -1 for different Temperatures [ o C] (for strips, the total noise is shown, for the pixels only the noise due to the leakage current is shown) Noise at end of Life per strip/pixel [e - ] for Temperature T 3000fb fb -1 R [cm] prerad T = -10 T = -20 T = -30 T = -10 T = -20 T = (12cm) (12cm) Roadmap for the SSD R&D Technology The purpose of this programme is to further investigate the very encouraging results obtained with miniature microstrip detectors made on p-type substrates and irradiated to SLHC fluences [19], shown in Fig. 3. These results were obtained with a few p-type substrate miniature detectors (1 cm 2 ) both oxygen-enriched and standard. A programme to study the possible benefit of high oxygen content in the n-type and p-type silicon in term of charge collection efficiency after irradiation is under way in RD50, and we will be part of this programme. The irradiation programme foreseen for the RD50 studies includes neutron as well as proton irradiation to investigate possible different degradation of the charge collection properties, in a similar way to the differences that were found in term of deterioration of the full depletion voltage with proton or neutron fluences [21]. As described above, the ATLAS tracker upgrade programme will require that the different options: (DO)FZ or MCz, p-type or n-type be studied for charge collection efficiencies with LHC speed electronics after doses up to and above the design requirements. The RD50 mask set only gives miniature test sensors, but with sense element dimensions close to ATLAS requirements, so this should be adequate for a first determination of any strong advantages for a particular technology in terms of radiation tolerance and operating requirements after irradiation. The following is a list of issues to be resolved by the Technology R&D program, before a prototype run is started. Preferably, the issues should be decided before the demonstrator run. a. Wafer type: FZ or MCz, p-type or n-type (Charge collection efficiency (CCE) vs. fluence, annealing of CCE) b. Strip isolation on p-type sensors: p-spray / p-implant /combined p-spray & p-implant (pre-and post-rad measurements of breakdown voltage, Cint, Rint) c. DC coupled SSD / AC coupled SSD (Paper study of cost savings vs. complexity and performance of front-end ASICs) Demonstrator Short Strip Sensors and Mechanical/Dummy Detectors The micro-strip sensors, to the desired performance, need to be manufactured with at least one supplier capable of producing the full order to establish the likely yield, possible cost drivers and prove that the radiation hardness required can be achieved in a large scale production.

11 ATLAS Project Document No: Page: 11 of 13 The SSD design need to be based on the best information for the final module design. The sensors will be used in prototyping module assemblies and for a limited campaign of radiation tests similar to that for the miniature detectors but with adapted mechanics (as used in the ATLAS SCT). The number of possible substrate technologies (FZ and MCz p-type) will depend on the outcome of the R&D programme based on the miniature detectors from the RD50 mask set (see above) and the Japanese program with HPK [18]. An important part of the prototyping procedure for any super-module or multi-module tests will be the ability to build large mechanical prototypes with as much realism as possible. These dummy detectors are inexpensive per item, the prototyping can require many hundred such objects. Once masks for the final processing are defined they can be easily designed but the processing costs, using appropriate silicon to the required thickness, can still be nonnegligible. Such dummy devices do not have to be produced with the final large volume supplier but do require the ability to handle, process and dice 300 µm thick 150 mm wafers Electrically Active Short Strip Sensors The final goal of this programme should be to generate prototype sensors with at least one viable large volume manufacturer to a design that would be useable for the ATLAS SLHC upgrade tracker. Such sensors are vital to proving the full functionality of any prototype module or super-module and are needed in sufficient quantity to qualify the manufacturer for any final orders to be placed. An essential aspect of this is an irradiation programme with adequate statistics at different doses for sensors tested with minimum ionising particles and read out with appropriate speed electronics. Sensor orders need to be placed relatively early due to the long delivery time, the requirements of acceptance testing by the collaboration and the need to start module production early given the experience needed before sites can enter into routine production. Having the sensor solution in place by the end of the decade is vital to achieving the current aggressive schedule for SLHC tracker construction. ATLAS will need to purchase sufficient sensors for its prototyping programme at many potential locations. This pre-production order will also need to form the subject of a major irradiation campaign which should include tests with full-readout connected to the sensors as well as the single-chip test-system Electrically Active Long Strip Sensors The purchase of 9 cm or 12 cm length sensors to the requirements of the outer radius module programme should proceed in parallel with that for the short strips. Studies using miniature detectors and short strip detectors irradiated to intermediate doses could help determine the long strip designs. A key issue will be the choice of n or p substrate for which cost will be a major driver. The issue of AC or DC coupling, given that shot-noise may not be dominant, should also be revisited from the perspective of cost. It is likely that short strip mask sets will be used for both n and p substrate sensors and that these can help determine the likely performance of 9cm or 12cm length sensors after neq/cm 2. Nevertheless, the requirements of an early start to module production for the much larger area of the high radii modules suggests that the programme to select the appropriate technology needs to be accelerated with respect to that for the short strips and that pre-production orders will be needed sooner than for the short strips Electrically Active Wedge Detectors It is likely that many mask sets will be needed for the forward discs making prototyping very expensive. Nevertheless, prototype modules will need to be proven and this programme cannot trail far behind that for the barrel modules Electrically Active Pixel/Large Pixel/Very Short Strip Detectors This programme is less clear due to current uncertainties in the Layout Group as to the appropriate radii to make the transition from pixels to strips and the possibilities of using an intermediate technology for one or more layers. For the pixel layers the assumption is that sensors to current technologies but with smaller pixel sizes should be ordered to fit with the availability of new deep sub-micron pixel electronics. The choice of the 12 6cm large area short-strip sensor does leave room for some small prototypes on the same wafer. The b-layer will probably remain the subject of intense R&D with mainly small test detectors until fairly late in the programme.

12 ATLAS Project Document No: Page: 12 of Cost The following is a rough estimate of the costs of this program a. Technology development: Assuming that the existing mask sets (in Japan and from RD50) can be used, and that FZ and MCz are to be studied in both n and p type within the collaboration, we need processing of about 5 wafers (6 ) of each of the four substrate types. This programme would be expected to cost at least CHF 50k. b. Demonstrator Short strip Sensor This requires a new mask set, and processing would be done for two wafer types maximum. We would need about 100 sensors, i.e. of the order 12 wafers. This would require about CHF 100k. The required number of mechanical dummies would be about 200, with a cost of about CHF 30k. c. Prototype electrically active short sensors This requires a new mask set, and processing would be done for one wafer type. We would need about 200 sensors, i.e. of the order 25 wafers. This would require about CHF 150k. d. Prototype electrically active long sensors This requires a new mask set, and processing would be done for one wafer type. We would need about 50 sensors, i.e. of the order 50 wafers. This would require about CHF 200k. e. Prototype electrically active wedge sensors This requires a new mask set, and processing would be done for one wafer type. We would need about 50 sensors, i.e. of the order 50 wafers. This would require about CHF 200k. 5.6 Summary, Including Milestones The aim of the short strip programme is to demonstrate large area sensors of the required granularity for an instantaneous luminosity of cm -2 s -1 and able to withstand radiation doses equivalent to roughly 1x10 15 neq/cm 2 can be provided in sufficient quantities, affordably to ATLAS for the SLHC. To this end the following milestones are proposed. The extend to which the milestones can be achieved will depends on the availability of funds. : Winter Orders placed with R&D suppliers for RD50 multi-project wafers for ATLAS programme Spring Mask designs completed and orders placed for large area FZ p-type demonstrator short strip sensors manufactured by a company capable of large volume production. Masks to include large pixel/very short strip test detectors. Spring2007. Delivery of diced RD50 wafers with miniature sensors relevant to ATLAS programmes Summer Delivery of FZ p-type demonstrator short strip sensors and large pixel/very short strip test detectors manufactured by company capable of large volume production. Fall Delivery of dummy sensors for mechanical studies. (Could include rejects from FZ p-type run.) Winter Completion of irradiation programme with RD50 miniature detectors relevant to ATLAS technology choices (pixels and strips) and irradiation of several of the large area FZ p-type demonstrator short strip and large pixel/very short strip sensors. Spring 2008 Decision on optimal long strip substrate technology compatible with large volume production Spring Decision on optimal short strip substrate technology compatible with large volume production Spring Mask design completed for large area ATLAS prototype long strip wafers and order placed Summer Mask design completed for large area ATLAS protottype short strip wafers and order placed Summer2008. Pixel sensor mask design for reduced pixel dimensions completed and order placed Summer Sensor mask design for large pixel/very short strip completed and order placed Fall First diced prototype long strip sensors delivered Winter First diced prototype short strip sensors delivered Winter First reduced size pixel pre-series sensors delivered Winter First large pixel/very short strip pre-series sensors delivered Spring Pre-series sensors tested and delivered for prototype module construction

13 ATLAS Project Document No: Page: 13 of 13 Summer First irradiation results with all pre-series sensors Appendix A