1. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 1 Bolek Wyslouch Massachusetts Institute of Technology for the CMS Collaboration Winter Workshop Big Sky, MT February 2009 CMS Experiment at LHC: Detector Status and Physics Capabilities in Heavy Ion Collisions CMS HI groups: Athens, Auckland, Budapest, CERN, Colorado, Cukurova, Ioannina, Iowa, Kansas, Korea, Lisbon, Los Alamos, Lyon, Maryland, Minnesota, MIT, Moscow, Mumbai, Paris, Seoul, Vanderbilt, UC Davis, UI Chicago, Zagreb c d a b

2. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 2 Heavy Ion Physics at the LHC Pb+Pb Collisions at  s NN ~ 5.5TeV Large Cross section for Hard Probes High luminosity 10 27 /cm 2 s Copious production of high p T particles Nuclear modification factors R AA out to very high p T Large cross section for J/ψ and  family production  cc LHC ~ 10x  cc RHIC  bb LHC ~ 100x  bb RHIC Different “melting” for members of  family with temperature Large jet cross section Jets directly identifiable Study in medium modifications

3. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 3 Kinematics at the LHC Access to widest range of Q 2 and x J/  Z0Z0  Saturation Coverage Rate/Trigger

4. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 4 CMS detector at the LHC MUON BARREL Drift Tube Chambers ( DT ) Resistive Plate Chambers ( RPC ) SUPERCONDUCTING COIL IRON YOKE Silicon Microstrips Si Pixels TRACKER ECAL Scintillating PbWO4 crystals HCAL Plastic scintillator/brass sandwich human Length: 21.6 m Diameter: 15 m Weight: ~12500 tons Magnetic Field: 4 Tesla Cathode Strip Chambers (CSC )‏ Resistive Plate Chambers (RPC‏ MUON ENDCAPS

5. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 5 CASTOR CASTOR ZDC ZDC TOTEM RPs TOTEM T2 CMS (+TOTEM): Largest phase-space coverage ever in a collider. -420m (FP420)‏ 420m CMS detector will extend deep into LHC  ~14 ZDC funded by US DOE NP & NSF

6. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 6 Particle detection layers in CMS

7. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 7 LHC proton beam on September 10th CMS recorded several splash events when the proton beam was steered onto the collimators; halo muons were observed once beam started passing through CMS Detectors flooded by few hundred thousand muons resulting from 10 9 protons (one bunch) simultaneous hitting a collimator Correlation between the energies collected in the HCAL and ECAL calorimeters from several “beam-collimator collisions”

8. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 8 Cosmic Muons in CMS A cosmic muon that traversed the barrel muon systems, the barrel calorimeters, and the silicon strip and pixel trackers Improved track quality in the strip tracker as the nominal design geometry is replaced by versions aligned with cosmic muons

9. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 9 Tracker Performance  On track Strip clusters S/N ratio, corrected for the track angle  TOB thick sensors: S/N = 32  TIB/TID thin sensors: S/N = 27/25  TEC (mixed thickness): S/N = 30  Track hit finding efficiency  TIB and TOB layers  Muon momentum distribution  high quality tracks (8 hits, one in TIB layers 1-2, one in TOB layers 5-6)  Partial CRAFT statistics (expected >70K tracks at P T >100 GeV for full CRAFT) S/N Entries Layer

10. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 10 Physics cases for HI@CMS Parton density, transport coefficient Jet “quenching”6 ε crit, T crit Quarkonia suppression5 Thermodynamics & transport properties Hard-probes (triggering)4 Hydrodynamics, Medium viscosity Elliptic Flow3 Hydrodynamics, Equation of State Low p T π/K/p spectra2 Color Glass Condensate, xG A (x,Q 2 ) dN ch /dη1 Global event characterizationMB L1 trigger, centrality0 in order to learn about…We will look into…Case no. ^

11. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 11 pp Minimum Bias using Forward HCAL 3<|  |<5 HF   Min-bias trigger: Number of HF trigger towers over threshold > N: any side (OR), or on both sides (AND) of I.P. Inverse weighting OR all evts Unfolding the effect of HF min-bias triggering on physical measurements Re-weighted charged-particle multiplicity per evt. preliminary AND

12. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 12 Charged-Hadron Spectra pp One of the first measurements to be done: charged-hadron spectrum Important tool for calibration, alignment & understanding of detector response Min-bias and/or Zero-bias trigger Statistics: ~ 2 million events d 2 N/dydp T  Simulation “Data” pp kk  CMS Si-Strip 0.1 1 10 p T (GeV)  kk pp CMS Si-Pixel 0.1 1 10 p T (GeV) dE/dx (MeV/cm) KK pp 0 1 2 p T (GeV)

13. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 13 Efficiency o Fake Rate CDF Transverse Longitudinal Pb+Pb, dN/dy = 3500 World’s best momentum resolution Excellent impact parameter resolution Pb+Pb, dN/dy = 3500 CMS Tracking Performance PbPb Occupancy in Pb+Pb dN/dy = 3500 1-2% in Pixel Layers < 10% in outer Strip Layers Efficiency ~75% above 1 GeV/c

14. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 14 High Level Trigger Pb+Pbp+p Input event rate3kHz (8kHz peak)100kHz Output bandwidth225 MByte/sec Output rate10-100Hz150Hz Rejection97-99.7%99.85% Level-1 Pb+Pbp+p Collision rate3kHz (8kHz peak)1GHz Event rate3kHz (8kHz peak)32MHz Output bandwidth100 GByte/sec Rejectionnone99.7% CMS Trigger and DAQ in Pb+Pb vs p+p Level 1 trigger - Uses custom hardware - Muon tracks + calorimeter information - Decision after ~ 3 μsec High level Trigger - ~1500 Linux servers (~10k CPU cores) - Full event information available - Runs “offline” algorithms to be partially funded by US DOE NP

15. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 15 Rates to tape with HLT Min bias with HLT Min bias Hard Probes: HLT vs Min Bias J/  Y and Jet reconstruction available at HLT Example trigger table: HLT improves hard probe statistics by more than a factor of 10!

16. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 16  ● produced in 0.5 nb -1 ■ rec. if dN/d  ~ 2500 ○ rec. if dN/d  ~ 5000 curves from Nucl. Phys. B492 (1997) 301–337 Statistical reach Quarkonia:  →  Expected rec. quarkonia yields: J/  : ~ 180 000  : ~ 26 000 Detailed studies of Upsilon family feasible with HLT Statistical accuracy (with HLT) of expected  ’ /  ratio versus p T -> model killer...

17. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 17 Good Resolution Pb+Pb dN/dy = 5000 p+p High Efficiency Jet studies from CMS Note 2003/004 Heavy Ion Jet Finder Performance A modified iterative cone algorithm running on calorimeter data gives good performance in Pb Pb collisions Offline jet finder will run in the HLT Other jet algorithms are being studied

18. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 18 High p T Suppression Clear separation of different energy loss scenarios Efficient jet trigger is essential to extend the p T range

19. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 19 Jet E T reach Jet spectra up to E T ~ 500 GeV (Pb-Pb, 0.5 nb -1, HLT- triggered) Detailed studies of medium-modified (quenched) jet fragmentation functions

20. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 20 Event/Centrality selection Reaction plane determination Vertex finding Track reconstruction Jet finding Photon identification Ingredients: All results based on full GEANT-4 simulations using full reconstruction algorithms on expected one run-year statistics q,g Hadrons Jet axis q,g Photon Jet axis provides parton direction Photon energy tags parton energy E T Charged hadron tracks used to calculate z = p T /E T Multiplicity and flow measurements characterize density, path length Photon-tagged jet fragmentation functions Measure jet fragmentation function: dN/dξ with ξ=-log z Main advantage Photon unaffected by the medium Avoids measurement of absolute jet energy DOE Review Driven..

21. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 21 Photon-tagged fragmentation functions generator-level generator- level σ~13% Use isolated photons + “back-to-back” cut on azimuthal opening angle between the photon and the jet to suppress NLO and background events  Determines FF with <10% deviation LO NLO 172 o

22. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 22 Signal and background statistics Study for one nominal LHC Pb+Pb run “year” 10 6 sec, 0.5nb -1, 3.9 x 10 9 events Use 0-10% most central Pb+Pb dN/dη| η=0 ~2400 Simulate signal and background QCD (p+p) events Mix into simulated Pb+Pb events (~1000 events)‏ generator-level Total 122158 Total 37490

23. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 23 Reconstruction Tracking Low p T cutoff at 1GeV/c Efficiency (algorithmic + geometric) ~ 50-60% Fake rate ~ few % Jet finding Iterative cone algorithm with underlying event subtraction (R=0.5)‏ Performance studies on away-side jet finding Photon ID Reconstruction of high-E T isolated photons Jet finding: Eur. Phys. J. 50 (2007) 117 Tracking: NIM A566 (2006) 123

24. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 24 Before cuts:After cuts: S/B=4.5 S/B=0.3 WP=60% S eff Photon identification performance Quenched Pb+Pb E T   70GeV Photon isolation and shape cuts improve S/B by factor ~15

25. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 25 Fragmentation function ratio Medium modification of fragmentation functions can be measured High significance for 0.2 < ξ < 5 for both, E T   0GeV and E T   0GeV E T   70GeV E T   0GeV Reco quenched Pb+Pb / MC unquenched p+p

26. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 26 Few words about schedule Just announced by DG in a short e-mail, result of discussions at the annual “Chamonix” planning meeting LHC will get physics data at the end of 2009 There is a strong recommendation to run during winter Collect 200 pb -1 pp at 5 TeV/beam and finish it with the short HI run (10 TeV E cm pp ~ 4 TeV/nucleon PbPb) More details will be available in the next few weeks.

27. Bolek Wyslouch/MITBig Sky, 7 Feb, 2009 27 Summary The CMS detector has excellent capabilities to study the dense QCD matter produced in very high energy heavy-ion collisions, through the use of hard probes such as high-ET (fully reconstructed) jets and heavy quarkonia With a high granularity inner tracker (full silicon, analog readout), a state-of-the-art crystal ECAL, large acceptance muon stations, and a powerful DAQ & HLT system, CMS has the means to measure charged hadrons, jets, photons, electron pairs, dimuons, quarkonia, Z 0, etc ! The CMS Heavy Ion group is busily preparing for the upcoming physics runs in pp and PbPb