ATLAS Pixel Detector Upgrade: The Insertable B-Layer

UNIVERSITY OF OKLAHOMA GRADUATE COLLEGE HOMER L. DODGE DEPARTMENT OF PHYSICS AND ASTRONOMY ATLAS Pixel Detector Upgrade: The Insertable B-Layer A REPORT SUBMITTED TO THE GRADUATE FACULTY in partial fulfillment of the requirements for the SPECIALIST S EXAMINATION by DAVID BERTSCHE October 31, 2012

Abstract The Insertable B-Layer (IBL) is an upgrade to the ATLAS Pixel Detector with installation planned for mid 2013. In addition to increasing the lifespan of the detector, this upgrade will improve its capabilities. This paper discusses the physics motivation behind the pixel detector and how semiconductors are used in particle detectors. The design, manufacturing, installation and commissioning of the IBL are also discussed.

Contents 1 Introduction 1 2 Motivation: b Physics 1 2.1 Top Physics......................................... 2 2.2 Higgs Studies........................................ 2 3 Pixel Detector Design Requirements 2 3.1 LHC Accelerator...................................... 3 3.2 The ATLAS Detector................................... 4 3.3 Tracking and Vertexing.................................. 4 3.3.1 Tracking....................................... 5 3.3.2 Vertexing...................................... 5 3.4 b Tagging.......................................... 6 4 IBL Engineering 7 4.1 Solid State Detectors.................................... 7 4.2 Radiation Passage Through Semiconductors....................... 9 4.3 IBL Design Requirements................................. 9 4.4 Modules........................................... 10 4.4.1 Sensor........................................ 10 4.4.2 Readout Electronics................................ 10 4.5 Radiation Hardness..................................... 11 4.6 Layout............................................ 12 4.7 Supporting Hardware................................... 12 4.8 Manufacturing....................................... 12 4.9 Installation......................................... 14 5 Commissioning 14 6 Conclusion 15 A Appendix 16

1 Introduction The ATLAS 1 experiment is a large collaboration involving 174 institutions from 38 nations; it is one of four major detectors participating in the Large Hadron Collider (LHC) experiment. The detector contains many specialized subsystems including the Inner Detector. The ATLAS Inner Detector contains a Pixel Detector that specializes in identifying interaction vertices. The inner layer of this Pixel Detector, known as the B-layer, suffers so much radiation damage that the efficiency of the Pixel Detector deteriorates. After approximately five years of LHC operation, it is expected that the B-Layer will have received the maximum cumulative radiation dose for which it was designed. Figure 1: The Inner Detector is at the center of the ATLAS detector.[1] Rather than replacing the B-Layer, the solution is to insert a new layer of detector modules between the beam pipe and the present detector; this is the Insertable B-Layer (IBL). In addition to extending the lifespan of the detector, the upgrade will enhance its functionality. The Pixel Detector specializes in identifying interaction vertices; this is useful for many physics analyses. b physics, which is discussed in the following section, uses this technology. 2 Motivation: b Physics A chart of the particles in the Standard Model of Particle Physics (SM) is shown in Figure 3. The bottom (b) quark is an unstable, heavy, third generation fermion. 2 Any free quarks created by the LHC rapidly hadronize to form mesons 3 and baryons 4, and these soon fragment into a collimated spray of particles called a jet. If the jet progenitor is a particle containing a b quark then the jet 1 ATLAS: A Toroidal Lhc ApparatuS 2 Fermions are half integer spin particles. 3 Mesons are formed from a quark and an anti-quark. 4 Baryons are formed from three quarks. The proton and neutron are examples of baryons. 1

(a) The Pixel Detector is at the center of the Inner Detector. (b) The B -Layer is at the center of the Pixel Detector. Figure 2: Inner (a) and Pixel (b) Detectors [1] is called a b jet. The identification of a b jet is called b tagging. The IBL upgrade will enhance the detector s ability to b tag (technical details are discussed in section 3). Top physics and Higgs studies exhibit the usefulness of b tagging. 2.1 Top Physics Top quarks are very short lived, with a lifetime on the order of 10 25 s. Since this is an order of magnitude lower than hadronization timescales, this does not provide enough time to form top-flavored hadrons or tt -quarkonium-bound states.[2] The proton-proton collisions of the LHC most commonly produce top and anti-top pairs through gluon interactions (gg tt ). The most common decay products of the top quark are the bottom quark and an accompanying W boson, as shown in Figure 4. The SM predicts that this decay channel has a probability of nearly 100%. Experimental evidence of a BR(t W b)5 that deviates significantly from 1 could be evidence for physics beyond the SM.[3] Since the top quark is not detected directly, it must be studied through its decay products; thus a detector that can efficiently b tag aids top physics by removing a large amount of background. 2.2 Higgs Studies The H bb decay channel is one of the channels used in the Higgs Boson search. Even though this channel does not yet provide enough statistical significance for discovery, the Higgs is expected to couple preferentially to high mass particles such as the b quark. In fact, if the SM Higgs mass is indeed 125 GeV, H bb is the dominant decay mode (Figure 5). So this channel is expected to prove powerful in measuring the relative Higgs branching ratios and its spin.[4] 3 Pixel Detector Design Requirements This section gives a brief overview of some of the aspects of the LHC and ATLAS that are relevant to the IBL, with a focus on the IBL s importance for current particle physics research. 5 BR: Branching Ratio. The probability that a certain decay will occur relative to other possible decays. 2

Figure 3: The elementary particles of the Standard Model. 3.1 LHC Accelerator Since the ATLAS detector is part of the LHC experiment, the LHC design determines the luminosity delivered to each detector. Luminosity is a function of the number of particles per bunch and the collision rate. These parameters in turn determine the design parameters of the detectors, such as radiation hardness requirements and readout speed requirements. The LHC was constructed in the 27 kilometer long circular tunnel remaining after the Large Electron-Positron (LEP) accelerator was decommissioned in 2000. It is designed to accelerate both protons (p) and heavy ions (A), in particular lead nuclei.[5] Stronger, superconducting electromagnets were designed and installed in order to bend the heavier particles to a sufficient radius of curvature. The LHC is planned for an ultimate p-p energy of 14 TeV at a design luminosity of 10 34 cm 2 s 1 and an ultimate A-A energy of 5.5 TeV at a design luminosity of 10 27 cm 2 s 1. 6 The LHC is designed for counterrotating particle beams containing 2808 bunches of 1.15 10 11 particles per bunch that travel through two beam pipes. The beam pipes converge at the detectors allowing bunches to cross every 24.95 ns with average of 19.02 particle collisions per bunch crossing. These collision rates, energies and luminosities are unprecedented and required the development of faster and more radiation-hard electronics. 6 The maximum p-p energy reached as of writing this paper is 8 TeV at a peak luminosity of 10 33 cm 2 s 1. 3

Figure 4: t W b and subsequent decays Figure 5: Standard Model Higgs Boson branching ratios, including theoretical uncertainties.[2] 3.2 The ATLAS Detector ATLAS is a 4π particle detector, which means that it covers almost the entirety of the solid angle around the interaction point. Each subsystem of the detector consists of a barrel section and two end caps, as shown in Figure 1 and Figure 2, which does leave small cracks through which particles can escape undetected. ATLAS uses a right-handed coordinate system; the x-axis points towards the center of the LHC ring, the z-axis points along the beam pipe and the y-axis is vertical. The Inner Detector (ID) is comprised of the Transition Radiation Tracker (TRT), the Silicon Microstrip Tracker (SCT) and the Pixel Detector, as shown in Figure 6. The ID is operated in a 2 T solenoidal field in order to curve the trajectory of charged particles. The TRT and SMT are both used to track charged particles; the TRT is made of straw tubes containing a xenon-based gas and the SMT is made of silicon strips. The main function of the Pixel Detector is vertexing and measuring impact parameters (IP). 3.3 Tracking and Vertexing Tracking and vertexing performance are critical in exploiting the full power of the LHC. Pixel detectors have the high space resolution, time resolution and granularity needed to provide this performance. Simulations show that the IBL will improve primary vertex reconstruction RMS 4

Figure 6: Cut-away view of the Inner Detector.[4] resolution from 15 µm to 11 µm (in the x and y directions) and from 34 µm to 24 µm (in the z direction).[4] The gains in vertex and IP resolution significantly improve b tagging performance. 3.3.1 Tracking The tracking systems of the ATLAS detector establish the presence and trajectory of charged particles. These particles can originate from the interaction point 7 or from subsequent decays. Because a magnetic field permeates the tracking systems, these charged particles travel in a curved trajectory. The transverse momentum 8 of a charged particle can be determined from the radius of curvature of its path and charge can be determined from the direction of curvature. The IBL will improve tracking by improving the IP resolution (Figure 7). This is because it is closer to the beam and has modules with a smaller z pitch than the present B-Layer. However, since the overall track length will be nearly unchanged, there is only a small contribution to the measurement of track curvature. 3.3.2 Vertexing Vertexing means tracing tracks back to their origin and deducing the location of primary and secondary interaction vertices in three dimensional space. The primary vertex is the point where two accelerated particles (such as protons) in the beam pipe collided head-on and produced many particles. In Figure 7 a particle of interest is labeled the primary particle. The secondary vertex 7 The point two particles in the beam (e.g. protons) collide head-on. 8 Transverse momentum (p T ) is the projection of total momentum onto the x-y plane. 5

is the point where the primary particle decayed into two or more other particles. Identifying the vertex locations aids in measuring the transverse momenta of individual particles and in identifying the flavor 9 of jets. Figure 7: The Impact Parameter (IP) is determined from the relative geometry of the primary and secondary vertices. Tracks 1 and 2 represent decay products of the primary particle. 3.4 b Tagging Jets from up, down and strange quarks are called light jets and usually cannot be distinguished from each other. Jets from charm quarks are sometimes grouped with light jets and sometimes uniquely identified. The short lifetime of the top (t) quark means the secondary decay vertex cannot readily be distinguished from the primary interaction vertex and jets are not directly t tagged. In contrast, the b quark must cross generations to decay, and thus most mesons and baryons containing the b quark have a relatively long lifetime of 10 12 s. Thus these particles can travel a few millimeters from the primary vertex before decaying. Therefore b jets can be identified by a secondary vertex with a large positive IP. 10 In addition the presence of a soft 11 lepton with the appropriate relative transverse momentum can be used to b tag and can be particularly useful in distinguishing between b and c jets.[6] Computer algorithms are used to b tag; their parameters are adjusted with the goal of maximizing efficiency while minimizing the number of jets erroneously b tagged (i.e. mistagged). 12 These algorithms are tested and optimized using Monte Carlo simulations of detector events before being used on data from the experiment. Efficiency = # of jets that were b tagged actual # of b jets 9 The type of quark that the jet originated from. 10 A negative IP means the secondary vertex comes before the primary vertex with respect to the primary particle direction. This indicates that the detector resolution was not fine enough to accurately draw the particle tracks. 11 Low momentum, 5 GeV. 12 Typical efficiencies are 60%-70%, with a mistag rate of 1%. 6

4 IBL Engineering 4.1 Solid State Detectors Semiconducting material is used to detect ionizing radiation by operating it as a solid state ionization chamber. When ionizing radiation passes through a semiconductor, an electron in the crystal lattice can be excited from the valence band into the conduction band leaving a hole behind. This creates so-called electron-hole pairs. The electrons are then collected for preliminary processing and amplification into a signal that is sent off the detector for subsequent analysis. The basic solid state detector is essentially a silicon p-on-n diode operated with a reversed bias voltage. Pure silicon forms a hexagonal crystal structure with each atom having four valence electrons. Impurities can be added to the silicon crystal structure in a process called doping. If the dopant atom has 5 valence electrons then they are called donors (because they can donate an electron to the crystal). Silicon doped with donors is called n-type. If the dopant atom has 3 valence electrons then they act as acceptors (because they can accept an electron from the crystal). Silicon doped with acceptors is called p-type. When n-type and p-type silicon are brought together, some electrons and holes near the junction diffuse and combine creating a depleted junction region. This creates a net negative charge on the p side and a net positive charge on the n side leading to an electric field that prevents further diffusion. In the band model shown in Figure 8 this equilibrium condition can be represented by the alignment of the Fermi levels (E F ) of the two sides. Figure 8: n-type and p-type silicon in isolation (a) and after combination (b) into a diode. The central region around the junction has an absence of electrons or holes and is known as the depletion zone. E F is the Fermi level (potential energy) of the electron/hole, E C is the top of the conduction band, E V is the bottom of the valence band, E i is the intrinsic Fermi level that undoped silicon would have and V bi is the built in voltage produced. [7] Ionizing radiation can produce electron-hole pairs anywhere in the silicon sensor. However if they form outside of the depletion zone they will soon recombine. If they form in the depletion zone, the electric field present there will cause the electrons and holes to move in opposite directions. In order to increase the useful volume of the silicon sensor, a reverse bias voltage is applied by means of conducting electrodes so that the depletion zone grows. This applied voltage must be less than the breakdown voltage at which point the electric field is too high for the internal structure of the diode 7

and leakage current 13 is produced. This means the movable charge carriers are accelerated strongly enough by the external field that they collide with the lattice atoms creating more electron-hole pairs in an avalanche multiplication process. The same electrodes that apply the bias voltage are used to collect the electrons produced in the sensor. Figure 9 shows one cell of a pixel detector. The sensor bulk is divided into many pixels by using a pixel implant (as shown in Figure 10). The collected electrons are transferred to the electronics chip through a conducting bump bond.[8] The close proximity of the sensor and readout electronics allow for the high readout speeds required by ATLAS. Figure 9: One cell of a pixel detector.[8] Figure 10: A pixel detector module showing the sensor, bump bonds and readout electronics.[8] 13 Leakage currents contribute to noise and should be avoided, but small leakage currents are always present, even in the absence of radiation. Some additional sources are diffusion of charge carriers from the undepleted to the depleted volume and thermal generation of charge carriers at generation-recombination centers.[8] 8

(a) Rate of energy loss due to ionization as a function of kinetic energy of a charged pion traveling through silicon. The dotted line shows density and shell correction terms.[7] (b) Number of electron-hole pairs generated in a 300 µm thick sample of silicon.[8] Figure 11: The Bethe-Bloch formula applied to silicon. 4.2 Radiation Passage Through Semiconductors A charged particle traveling through matter experiences elastic collisions with the electrons in that material. This causes the charged particle to lose energy based on the distance traveled through the material. For moderately relativistic charged particles, other than electrons, this process is described by the Bethe-Bloch formula, [ de dx = 2πN orem 2 e c 2 ρ Zz2 Aβ 2 ln ( 2me γ 2 v 2 W max See Appendix A for an explanation of the variables. I 2 ) 2β 2 δ 2 C ]. Z As shown in Figure 11a, the amount of energy deposited by the charged particle has a minimum value. Most relativistic particles (such as muons) have a mean energy loss close to the minimum and are know as mip s (minimum ionizing particles).[2] The rise after the minimum is known as the relativistic rise and comes from rare large energy transfers to a few electrons. Approximately a third of the energy lost by a charged particle while traversing silicon goes into electron-hole production; the rest is eventually dissipated as thermal energy. 4.3 IBL Design Requirements The primary motivations for the B-Layer upgrade are to alleviate radiation damage and to provide improved measurement accuracy, with a long term motivation to provide the capacity to handle the increased luminosity associated with operating the LHC at increasingly higher energies and culminating with the installation of the High Luminosity LHC (HL-LHC) in 2020. The present B-Layer was designed for a lifetime integrated luminosity of 300 fb 1, and radiation damage will appear in it over time. This decreases the impact parameter resolution, which seriously deteriorates the efficiency of b tagging and vertexing. The IBL has a lifetime integrated luminosity of 550 fb 1 and is capable of restoring full b tagging functionality to the pixel detector even if the present B-Layer fails completely from radiation damage. The beam pipe section currently inside the 9

ID will be removed and replaced with one of a smaller radius. This will place the innermost layer of the detector closer to the interaction point, thus improving vertexing and track reconstruction. The present LHC is expected to achieve a peak luminosity of 2.2 10 34 cm 2 s 1, twice what the current Pixel Detector was designed for. The predicted maximum rate of 50 interactions per bunch crossing will lead to event pileup 14 and increased readout inefficiencies. The IBL is designed for a peak luminosity of 3 10 34 cm 2 s 1 and has a comparatively low occupancy 15 which will restore readout efficiencies even with increased luminosities. Better vertex resolution will also better distinguish between primary and secondary vertices.[4] The IBL project has some unique constraints that require the application of novel engineering solutions. The small size of the IBL introduces tight tolerances and clearances thus increasing the complexity and precision required in general. It also means the sensor elements can t be overlapped in the z direction and sensors with thin edges must be developed and used. Minimizing materials are critical to performance optimization and are being integrated into all elements of the IBL: a new thin sensor design, low density carbon foam staves, CO 2 evaporative cooling and aluminum electrical conduits. 4.4 Modules 4.4.1 Sensor Two sensor designs with a pixel size of 50 250 µm 2 were chosen (both of them developed out of the p-on-n sensor introduced above): planar thin edge n-in-n sensors for the low η 16 region and double-sided 3D thin edge sensors for the high η region where tracking and z-resolution will most benefit from the electrode orientation. Radiation damage will increase the full depletion voltage required to an expected value of more than 1000V. Since the maximum reverse bias voltage expected to be available is 700V, this means that under-depleted operation of the sensors will be necessary for a significant part of the detector lifetime. To cope with this, a planar n-in-n sensor design was chosen as shown in Figure 12. The double-sided processing allows for a guard ring concept that maintains the sensor edge at ground potential.[11] 3D sensors are very similar in design and operation to planar sensors except that the electrodes extend into the sensor bulk as shown in Figure 13. 4.4.2 Readout Electronics A new Front-End (FE) Integrated Circuit (IC), named the FE-I4 chip, has been developed with 130 nm feature size technology for use in the both the IBL and HL-LHC upgrades. The motivations for creating this chip are to decrease the power and material requirements while increasing radiation tolerance and data capacity associated with the higher hit rates expected. The overall chip size will be increased from the previous FE-I3 chip in order to reduce manufacturing costs, but the peripheral chip area will decrease so that the active fraction of the module grows from 74% to almost 90%. The hit data from each pixel are stored in an on-chip buffer until triggering is completed and then the data are either read out or dumped. The architecture of the FE-I4 chip has been designed 14 Pileup happens when data from multiple events overlap, because the pixels can t be read out and reset rapidly enough. 15 Occupancy is the fraction of pixels that register a hit for a given event. Low occupancy is desireable because this improves track reconstruction. 16 η = ln[tan( θ )] where θ is the angle in degrees measured from the beam (or z) direction. η (psuedorapidity) is 2 the zero mass approximation of rapidity, which is Lorentz invariant and thus used in preference to θ. 10

Figure 12: n-in-n sensor design. The p-n diode junction is near the bottom edge. The distance between the outer edge and the first pixel is decreased from 450 µm to 100 µm in a slim edge sensor design.[11] to exploit the prediction that 99.75% of the hits will not pass triggers so that data will not be read out.[9] Negative charges are collected from the sensor, and each pixel has independent amplification, discrimination and threshold adjustment. An externally supplied 40 MHz clock provides timing, and time over threshold measurements (ToT) are stored with 4-bit resolution (Figure 14). Data from a maximum of 256 clock cycles are stored in the buffer for retrieval by the trigger. Data from a triggered pixel and its four neighbors are retrieved to enhance tracking resolution. 4.5 Radiation Hardness Exposing solid state sensors to intense nuclear radiation can drastically change the basic material bulk properties. These damages include the displacement of lattice atoms and nucleus transmutation through neutron capture. These defects can act as recombination-generation centers (meaning they spontaneously absorb and emit charge carriers) which increases the reverse bias current. They can act as trapping centers (meaning they trap charge carriers and re-emit them after some delay) which reduces detection efficiency. And they can change the charge density in the depletion zone, thus requiring an increased bias voltage to restore full sensor sensitivity. This eventually leads to type inversion.[7] Some steps can be taken to prevent and remedy radiation damage in the Pixel Detector. The introduction of oxygen impurities into the sensor bulk increases its tolerance to damage caused by charged hadrons. Room temperature annealing for about a week allows displaced lattice atoms to naturally move back to their positions. Because irradiation of the sensor bulk will eventually change the effective doping concentration leading to type inversion (from n-type to p-type), an n-in-n sensor design is used. After type inversion, the junction moves to the top side of the sensor allowing operation even if the bulk can 11

Figure 13: Double-sided slim edge 3D sensor design. An un-etched distance (d) of 20 µm is left for mechanical integrity.[11] not be fully depleted. This process is illustrated in Figure 15. Thus the sensor can still be operated even if the voltage required for full depletion exceeds the breakdown voltage. The FE electronics are also made of semiconductor material and suffer similar damage to that described for the sensors. Radiation-hard electronics are produced by using high quality gate oxides, guard rings around individual transistors and silicon on insulator (SOI) techniques. 4.6 Layout Figure 16 is a schematic showing a quarter section of the new beam pipe and IBL. There presently exists a radial free space of 8.5 mm between the beam pipe and the Pixel Detector; in addition, the inner radius of the beam pipe will be reduced from 29 mm to 25 mm. This gives a radial space of 12.5 mm into which the IBL will be inserted. The modules will be mounted on 14 staves arranged cylindrically around the beam pipe. Full hermetic coverage in φ is possible by tilting the staves by about 27 compared to the radial direction. However full coverage in the z direction is not possible due to space constraints. The coverage gap in the z direction is minimized by using sensors with slim edges. 4.7 Supporting Hardware Cooling fluid (CO 2 ) will run through a cooling channel in each stave to carry away heat generated by the modules to maintain the IBL at close to the target temperature of -15 C. Low voltage supplies, high voltage supplies and optical data links are delivered to the modules through service cables. 17 4.8 Manufacturing A flip chip bump bonding process is used to integrate the FE-I4 chips and the sensors into a bare module. The sensor and readout chip wafers are thinned and diced before bump bonding through a 17 Copper-clad aluminum wires are used for the low voltage supply. Pure copper would be sturdier and more compact, but would increase the overall radiation length of the electrical service material. 12

Figure 14: Calculation of time over threshold(tot). If the amplified analog signal crosses the threshold value then a hit is registered. The ToT value measures the strength of the signal and is read out digitally.[10] Figure 15: Before type inversion (a) the electric field grows from the bottom and must reach up to the pixel implants for the sensor to be operational. After type inversion (b) the electric field grows from the top and allows operation even if the bulk is only partially depleted.[4] reflow 18 operation. A carrier wafer support element is used to prevent heat-induced bowing during bonding and is removed from the flip chip assembly during the final step. A flexible printed circuit board (flex) and wire bonds for external connections are then added to the bare module to create a flex module. The support staves have integrated cables into which the flex modules will directly connect; this technique reduces the amount of dead material in the detector compared to previous connection methods. A number of modules will be reserved for qualification and testing. Radiation tolerance testing will be done to ensure that hit efficiency and position resolution remains above the minimum required values after receiving expected levels of radiation. A burn-in program involving 50 thermal cycles between 40 C and -35 C with the power on will be carried out to verify that an acceptable fraction of channels remains live. A sample of wire bonds will be destructively pulled to track any change of bond strength throughout the assembly and test process. 18 Reflow refers to the process of first applying the bump bond metal in a patch then heating to 180 C. Surface tension causes the metal to form into a sphere. 13

Figure 16: Schematic showing an x-y cross-section of the IBL design.[4] 4.9 Installation During the planned LHC shut down in 2013 the ATLAS end caps will be removed (see Figure 1) to gain access to the ID. Before installing the IBL, a 7.3 m section of the existing beryllium 19 beam pipe will be removed. This involves disconnecting associated supports, flanges, collars and wires; special tooling has been developed for this delicate task. Each step will first be tested on a full scale mock-up of the Pixel Detector. In order to support the IBL during the insertion process an IBL Support Tube (IST) will first be inserted. The tooling to insert the IST must be able to remotely anchor it to the wire suspension system of the present beam pipe. 5 Commissioning Even though the IBL contains many unique components, it is designed to be a subsystem of the current Pixel Detector rather than a new sub-detector of its own. The technology and software that will be used to collect and process IBL data are shared with the Pixel Detector.[4] 19 Because of its low atomic number (4) beryllium is relatively transparent to ionizing radiation. 14

After installing the IBL, the various systems will be commissioned. For example, commissioning the cooling system involves verifying that the cooling fluid connections are within leakage tolerances, verifying the thermal properties (ability to dissipate heat) of the modules, and testing each part of the cooling system loop. It is important to verify that power cuts off automatically if the cooling system fails. Having power on without cooling could damage the IBL modules if the temperature exceeds 40 C. A calibration and tuning software suite will test the communication links, front-end electronics, sensors and modules. Basic checks verify functionality, such as the integrity of bump bonds, wire bonds and optical communication links. Calibration refers to measuring the basic functionality of the IBL. For example, leakage current is measured initially and then monitored throughout operation in order to keep track of radiation damage. Tuning refers to adjusting parameters to optimize functionality. For example, the discriminator threshold (the level at which a signal is considered a hit) can be set at a module by module and a pixel by pixel basis. It must be high enough to filter out noise but low enough to provide efficient hit detection. Before operation with the LHC beam, data will be taken from random noise triggers and cosmic rays. Random trigger data are useful for testing data flow through the complete readout chain. Cosmic ray data will be used to align the IBL with respect to the rest of ATLAS; this is valuable preparation for further tracker alignment with the LHC beam data. 6 Conclusion A vertex system capable of b tagging with a high degree of accuracy is essential for the physics goals of the ATLAS detector. Radiation damage seriously deteriorates the efficiency of the Pixel Detector s performance of this function. The IBL will restore full vertexing functionality to the Pixel Detector while also improving its resolution. Many novel engineering solutions have been developed for the design, manufacture and installation of the IBL. The current schedule plans for installation to take place in mid 2013. 15

A Appendix Bethe-Bloch formula is given by [ ( de dx = 2πN orem 2 e c 2 ρ Zz2 2me γ 2 v 2 ) W max Aβ 2 ln I 2 2β 2 δ 2 C ]. Z where 2πN o rem 2 e c 2 = 0.1535 MeVc 2 /g; x is the path length in g/cm 2 ; r e = e2 4πm ec = 2.817 10 13 cm and is the classical electron radius; 2 m e is the electron mass; N o = 6.022 10 23 mol 1 and is Avogadro s number; I is the effective ionization potential averaged over all electrons; Z is the atomic number of the medium; A is the atomic weight of the medium; ρ is the density of the medium; z is the charge of a traversing particle; β = v c, the velocity of the traversing particle in units of the speed of light; γ = 1 ; 1 β 2 δ is a density correction; C is a shell correction; and 2m W max = ec 2 βγ 2 1+2 me M 1+β 2 γ 2 +( me M )2 is the maximum energy transfer possible in a single collision.[7] 16

References [1] The ATLAS Collaboration, The ATLAS Experiment at the CERN Large Hadron Collider, JINST 3 S08003, (2008). [2] J. Beringer et al. (Particle Data Group), The Review of Particle Physics, Phys. Rev. D86, 010001 (2012). [3] R. Kehoe, M. Narain and A. Kumar, Review of Top Quark Physics Results, Int. J. Mod. Phys. A 23, 353 (2008). [4] The ATLAS Collaboration, ATLAS Insertable B-Layer Technical Design Report, ATLAS- TDR-19, (2010). [5] L. Evans and P. Bryant, LHC Machine, JINST 3, S08001 (2008). [6] G. Aad et al. [ATLAS Collaboration], Expected Performance of the ATLAS Experiment - Detector, Trigger and Physics, (2008). [7] G. Lutz, Semiconductor Radiation Detectors, Springer, Germany, (1999). [8] L. Rossi et al., Pixel Detectors, Springer, The Netherlands, (2006). [9] M. Karagounis et al., Development of the ATLAS FE-I4 pixel readout IC for b-layer Upgrade and Super-LHC, TWEPP proceedings, (2008). [10] M. Lindner et al., Comparison of hybrid pixel detectors with Si and GaAs sensors, Nuclear Inst. and Methods in Phys. Research A 466, 1 (2001). [11] T. Rohe, Sensor Concepts for Pixel Detectors in High Energy Physics., www.slac.stanford.edu/econf/c020909/trpaper.pdf