ATLAS Phase II For the High Luminosity LHC

ATLAS Phase II For the High Luminosity LHC IPRD 13 – Sienna Italy 07/10/2013 Dr. B. Todd Huffman on behalf of the ATLAS Collaboration ( Oxford University, United Kingdom ) 07/10/2013 B. Todd Huffman 1

CERN, 4 July 2012 Ladies and gentlemen, I think we’ve got it! Discovery of a Higgs-like particle coupling to gauge bosons 07/10/2013 B. Todd Huffman 2

Precision measurements of Higgs couplings Final states targeted to measure couplings (that have low signal rate at LHC): tt. H (with H γγ) • Allows precise measurement of top-Yukawa coupling • Cleanest signal (w. r. t WH/ZH) S/B ~20% • S/√B ~6 with 3000 fb-1 (x 2 better than 300 fb-1) 07/10/2013 B. Todd Huffman 3

Physics at HL-LHC • Is this “Higgs” really THE Higgs? ? • Also rare decays of known states (like top quarks) • Energy upgrade imminent!! – New states of matter to be found? • SUSY, Hidden SUSY, Z-prime, etc… • Highly exciting time! 07/10/2013 B. Todd Huffman 4

Outline High-Luminosity • Detector challenges – Radiation damage – Background rates • Tracking – Rad. studies; choice of detector technology – Detector design concepts (baseline) • Trigger – HL-LHC studies on electrons and muons – Tracking ROI trigger • Conclusions 07/10/2013 B. Todd Huffman 5

What we mean by “Phase 2” Upgrade schedule 2009 ~25 fb-1 2010 2011 LHC startup, √s= 900 Ge. V √s=7~8 Te. V, L=6× 1033 cm-2 s-1, bunch spacing 50 ns 2012 2013 2014 LS 1 >=75 fb-1 2015 2016 2017 2018 LS 2 >=300 fb-1 2019 2020 2021 2022 2023 LS 3 … 07/10/2013 2030? ~3000 fb-1 2024 Go to design energy, nominal luminosity √s = 13~14 Te. V, L ~ 1× 1034 cm-2 s-1, bunch spacing 25 ns Phase-0 Injector and LHC Phase-1 upgrade to full luminosity √s = 14 Te. V, L ~ 2× 1034 cm-2 s-1, bunch spacing 25 ns Phase-I HL-LHC Phase-2 upgrade, IR, crab cavities √s = 14 Te. V, L = 5× 1034 cm-2 s-1, luminosity leveling Phase-II B. Todd Huffman 6

Detector Challenges Peak luminosity (leveled) 1 to 5 x 1034 cm-2 s-1; 3000 fb-1 • Higher trigger rate need improved triggers rather than simply raising thresholds globally Baseline of the future Inner Detector traversed by an event with 230 Pile Up Multiple interactions per crossing <140> • Higher detector occupancy • Increasing reconstruction complexity Increasing fluences >1016 neq/cm 2 close to the beam pipe • Increased radiation damage • Increased activation of materials Aging electronics (obsolete technology) 7

Phase-II: 2021/2022 (LS 3) 18 month shutdown ATLAS detector upgrade • Replacement of the entire Inner Detector • LAr and Tile calorimeter electronics upgrades • Possible upgrade of Forward Calorimeters • Upgrade of Muon system • • Muon Barrel and Large Wheel trigger electronics • Possible upgrades of TGCs in Inner Big Wheels • Coping with a track trigger Forward detector upgrade Target Absorber Secondaries (TAS) and shielding upgrade TDAQ upgrade Software and computing Various infrastructure upgrades Common activities (installation, safety, …) Phase-II Lo. I: https: //cds. cern. ch/record/1502664? ln=en 8

The Detector Challenges roughly Split into Two Parts Ø <1 m radius Radiation Damage to components; ITK expected fluence at 14 Te. V (3000 fb-1 ) Ø >1 m radius Pile-up & Trig. Rates; all of the detectors but will show upgrades for Muons and Electrons 07/10/2013 ATLAS Inner Tracker (ITK) region 9

RD 50 Sensor Rad. Damage Studies Unannealed Neutrons 900 V Unannealed 26 Me. V Protons All studied n-strip readout substrates become more and more similar with irradiation. This is true after neutron, proton and pion irradiations and with Hamamatsu and Micron devices. Micron Neutrons: A. Affolder, et. al. , Nucl. Instr. Meth. A, Vol. 612 (2010), 470 -473. Micron 26 Me. V Protons: A. Affolder, et. al. , Nucl. Instr. Meth. A, Vol. 623 (2010), 177 -179. HPK Neutrons: K. Hara, et. at. , Nucl. Inst. Meth. A, Vol. 636 (2011) S 83 -S 89. HPK 26 Me. V Protons: New and unpublished A. Affolder – VERTEX 2011 10

All Silicon Inner Tracker (ITK) • n-in-p advantages – Single-sided process (less expensive) • n+-in-n detectors – Double-sided (more expensive) – Guard rings @ ground near amplifiers • Both can work at HL-LHC rad. Levels – (If carefully designed)… – (And if they are kept cold ~-20 o C) 07/10/2013 B. Todd Huffman 11

All Silicon Inner Tracker for HL-LHC Classical layout with barrel cylinders and endcap disks – “Utopia” establishes baseline performance and cost • no special triggering layers Strips (74 M channels) long strip layers stub cylinder solenoid coil at y cr t os ll wa • 5 barrel layers + 7 fwd disks • stub layer for overlap region • 2 Si sensors at 40 mrad stereo angle Pixels (638 M channels) short strip layers IP beam pipe pixel layers total of 14 hits with full coverage to η=2. 5 • Pixels to η<2. 7 (forward muon ID) • 4 barrel layers + 6 fwd disks • inner 2 layers replaceable: 25μm x 150μm • outer Pixel: 50μm x 150μm • sensors bump bonded to readout chip using 65 nm CMOS technology minimize gaps in coverage • • last strip disk at z=3 m, last pixel layer at 25 -30 cm (improve double track resolution) 12 small “stub” layer in barrel

Est. Hit occ. (Everywhere < 1%) Hit Occ. in % Simulations indicate no problems with pattern rec. at these levels. (note: 200 events pile-up for this study) 07/10/2013 B. Todd Huffman 13

• Outer pixel layers Pixel staves – About 1. 4 m long and 5 mm thick – Modules on both sides, overlap for full coverage, makes module mounting easier – n-in-p sensors (less costly) • Inner pixel layers – I-beam design linking neighbouring layers; Clamshell construction – Optimizes stiffness – n+-in-n sensors

Phase 2: Barrel Strip-Tracker End of Stave card Strip barrel detector 5 barrel layers, 3 x short strips (23. 8 mm) and 2 x long strips (47. 8 mm) Strip Pitch – 74. 5 mm Stave-concept construction Slide in – more reliable installation Fully incorporated det. Services 07/10/2013 B. Todd Huffman Need ~20000 of these …. 15

Concept: Tracker elements To create integrated, fully functional objects, which can be Barrel strip stave – Produced in parallel – Tested fully early in the assembly – Single staves are of limited value and loss of small number has small impact on project → Project robustness Outer pixel stave Pixel disk EC strip petal 16

Radiation tracker components • Optical data link – 4. 8 Gbps – “Versatile link” In. Ga. As • Pixels Micro-cables to escape highest Rad. zones – ~4 m along the beam line – Then switch to optical readout • Strips Versatile Link 07/10/2013 B. Todd Huffman Ga. As 17

Inner Tracker Summary • Rad. Damage Studies show good performance for n-implant silicon detectors. – Cost considerations mainly driving decision to use nin-p for Strips and outer pixels • Tracking coverage and hit occ. maintained. • Novel support structures under design – Stave concept – Cooling requirements mean Services (cooling, monitoring, control) incorporated into support structure. • Rad-hard and SEU tolerant Gbps readout systems needed 07/10/2013 B. Todd Huffman 18

Part II: Increased Trigger rates • L 0 added @ 500 KHz rate • L 1 moves to 200 KHz rate • Important!: Maintain 20 Ge. V threshold muons (sharpen it up) and elec. (add tracks) • ROI seeded Tracks at L 1, regional triggers • ROI = Region of Interest • Incorporated in FE electronics chips • Leads to trigger and electronics upgrades like muon system (but most sub-detectors need some upgrades) 07/10/2013 B. Todd Huffman 19

Trigger Evolution – Phase-II • L 0 : 500 k. Hz o/p rate • L 1 : 200 k. Hz o/p rate – Addition of Track information (L 1 Track) in Regions of Interest (Ro. I) 20

Muon Triggers Phase-1: • Additional Thin Gap Chamber (TGC) doublets (EIL 4) • Include information from Tile Extended Barrel Phase-2: • New Small Wheel (NSW): Vector tracking based on s. TGC and Drift tubes • Reject b. g. from n & g • Reduce fake muons Rates for 20 Ge. V p. T threshold @ 3 x 1034 cm-2 s-1: • No change: 50 k. Hz • All Phase-1 upgrades except NSW : 30 k. Hz • Adding NSW: 13 k. Hz 21

• • L 1 Track Maintain single lepton trigger thresholds at ~20 Ge. V by adding track information at L 1 ~factor 5 rejection with 95% efficiency for offline selected events w. r. t. no L 1 track case Muon Trigger MU 20: require track p. T>15 Ge. V in DR<0. 15 Electron trigger: Require track in DR<0. 15 0. 67< E/p < 1. 5 => Factor ~10 rate reduction 22

Sharpening Trigger rates 1/E 6 90% efficient ¤ Shown is a simple Power-law falling spectrum – Backgrounds fall faster (Red-Dashed) ¤ Idealized Trigger turn-on is made sharper in right-hand case (Solid Green) ¤ Events that Pass Trigger – (Solid Blue) ¤ ~4. 5 times reduction in rate in right-hand case ¤ In actual muon trigger, expect a factor of two reduction in rate. 07/10/2013 B. Todd Huffman 23

CONCLUSIONS v Higgs Discovery motivates luminosity upgrade of LHC. v Proposed machine upgrade for Phase II (circa 2022) presents great challenges for the ATLAS detector. v. Direct Radiation Damage and SEU’s v. Increased backgrounds (pile-up events) v. Shown how these are addressed in the Inner Tracker and the lepton triggers (obviously much more work is taking place in other sub-systems in parallel). v. We should have an excellent detector during the next decade’s exciting discoveries! THANK YOU. 07/10/2013 B. Todd Huffman 24

BONUS MATERIAL 07/10/2013 B. Todd Huffman 25

Strip staves 26

Strip forward Detector • 7 Disks • Different types of modules – “Petal” concept Petalet 07/10/2013 • Continue tracking coverage |h|<2. 7 • “Petalet” sub-unit under construction. B. Todd Huffman 27

Stave Concept – Barrel strip stave insertion and locking mechanics Single-edge Mounting scheme Staves “slide in” from end of barrel Running theme Throughout inner tracker – Services incorporated in support structures. 07/10/2013 B. Todd Huffman 28

HL-LHC Fluences at z=0 07/10/2013 B. Todd Huffman 29

Strip staves 30

Strips electronics & readout (prototypes – close packed text: G. Viehhauser) – ABCn 130: binary readout architecture (like SCT) but new protocol, 256 inputs for smaller hybrids, ROI and fast L 1 trigger block – HCC: interface and module controller (1 per hybrid) • LV Powering: either serial (SP) or DC-DC at each hybrid/module Hybrids • Sensors: n-in-p single sided design, 98 x 98 mm 2, 500 V Max • Hybrids: glued onto sensor • ASICs: a 130 nm CMOS chipset – Additional powering and protection chipset, prototyped and new versions in development • Readout is being tested using stavelets (goal: good noise performance) DC-DC converter board Example: DC-DC powered stavelet 4 modules 8 hybrids 160 ABCn 20 k channels 31

AT LA S Str ip Rea d. Out (Barrel and Forwar d) On-Detector GBTx HCC HCC VTRx Custom Rad-hard Off-Detector: COTS Optical engines: TX: Laser driver + laser arrays RX: p-i-n array + TIA/discriminator GBTx functionality in FPGA 32

Radiation tracker components Responsivity of Photodiodes: Tough power budget decisions Makes Optical links unattractive choice at Pixel radii. 07/10/2013 Giga. Bit Transmitter (GBTX) Custom chip Multiplexer w. Forward Error correction (for Single Event Upset mitigation) SEU tests show they come in bursts. Fe. C can correct up to 16 bits in a row. Data scrambled (helps DC balance) B. Todd Huffman 33

Why Upgrade? Physics programme at LHC only begun with √s= 7 -8 Te. V collisions After 4 th July 2012… Higgs boson precision measurements • Expected uncertainties on signal strength reduced by a factor of 2 -3 with HL-LHC • Ratio of partial widths to measure ratios of couplings and probe new physics at 5 -15% level Higgs self-coupling in SM becomes accessible only at HL-LHC luminosity Probing new Physics • SUSY and other New Physics beyond SM • Enhancements in vector boson scattering amplitudes • Rare processes such as FCNC decays of top accessible to 10 -5 Physics case: European Strategy Meeting (Sept. 2012, Kracow) http: //indico. cern. ch/conference. Display. py? conf. Id=182232