1 Jets and High-pt Physics with ALICE at the LHC Andreas Morsch CERN

2 Outline Introduction Jets at RHIC and LHC: New perspectives and challenges High-p T di-hadron correlations Reconstructed Jets Jet Structure Observables  -Jet Correlations

3 Jets in nucleus-nucleus collisions Jets are the manifestation of high-p T partons produced in a hard collisions in the initial state of the nucleus-nucleus collision. These partons undergo multiple interaction inside the collision region prior to fragmentation and hadronisation. In particular they loose energy through medium induced gluon radiation and this so called “jet quenching” has been suggested to behave very differently in cold nuclear matter and in QGP. The properties of the QGP can be studied through modification of the fragmentation behavior Hadron suppression Jet structure.

4 Jet Physics at RHIC In central Au-Au collisions standard jet reconstruction algorithms fail due to the large energy from the underlying event (125 GeV in R< 0.7) and the relatively low accessible jet energies (< 20 GeV). Use leading particles as a probe.  s = 200 GeV STAR  s NN = 200 GeV

5 Quantities studied p T (trig) p T (assoc) Hadron Suppression Similar R CP : Ratio central to peripheral Hadron Correlations: p T (trig) – p T (assoc)  (trig, assoc) … “away side” “same side”

6 Evidence for Jet Quenching In central Au+Au Strong suppression of inclusive hadron yield in Au-Au collisions Disappearance of away-side jet No suppression in d+Au Hence suppression is final state effect. Phys. Rev. Lett. 91, (2003). Pedestal&flow subtracted STAR

7 Surface emission bias RHIC measurements are consistent with pQCD-based energy loss simulations. However, they provide only a lower bound to the initial color charge density. Eskola et al., hep-ph/ R AA ~ for broad range of q

8 Jet Physics at LHC: Motivation Study of reconstructed jets increases sensitivity to medium parameters by reducing Trigger bias Surface bias Using reconstructed jets to study Modification of the leading hadron Additional hadrons from gluon radiation Transverse heating. From toy model  = ln(E jet /p hadron ) Reconstructed Jet  s = 5500 GeV A. Dainese, C. Loizides, G. Paic

9 Jet Physics at LHC: New perspectives E T >N jets 50 GeV 2.0  GeV 1.1  GeV 1.6  GeV 4.0  10 4 Jet rates are high at energies at which they can be reconstructed over the large background from the underlying event. Reach to about 200 GeV Provides lever arm to measure the energy dependence of the medium induced energy loss 10 4 jets needed to study fragmentation function in the z > 0.8 region. Pb-Pb O (10 3 ) un-triggered (ALICE) => Need Trigger

10 Jet Physics at LHC: New challenges More than one jet E T > 20 GeV per event More than one particle p T > 7 GeV per event 1.5 TeV in cone of R =  2 +  2 < 1 ! We want to measure modification of leading hadron and the hadrons from the radiated energy. Small S/B where the effect of the radiated energy should be visible: Low z Low j T Large distance from the jet axis Low S/B in this region is a challenge !

11 New Challenges for ALICE Existing TPC+ITS+PID |  | < 0.9 Excellent momentum resolution up to 100 GeV Tracking down to 100 MeV Excellent Particle ID New: EMCAL Pb-scintillator Energy resolution ~15%/√E Energy from neutral particles Trigger capabilities central Pb–Pb pp

12 ALICE Set-up HMPID Muon Arm TRD PHOS PMD ITS TOF TPC Size: 16 x 26 meters Weight: 10,000 tons

13 Di-hadron Correlations: from RHIC to LHC Di-hadron correlations will be studied at LHC in an energy region where full jet reconstruction is not possible (E < 30 GeV). What will be different at LHC ? Number of hadrons/event (P) large Leads to increased signal and background at LHC Background dominates, significance independent of multiplicity Increased width of the away-side peak (NLO) Wider  -correlation (loss of acceptance for fixed  -widow) Power law behavior d  /dp T ~ 1/p T n with n = 8 at RHIC and n = 4 at LHC Changes the trigger bias on parton energy PYTHIA 6.2 See also, K. Filimonov, J.Phys.G31:S513-S520 (2005)

14 Scaling From RHIC to LHC S/B and significance for away-side correlations Scale rates between RHIC and LHC Ratio of inclusive hadron cross-section N(p T ) ~ p T 4 p T trig > 8 GeV RHIC/STAR-like central Au-Au ( events) LHC/ALICE central Pb-Pb (10 7 events), no-quenching From STAR p T trig = 8 GeV/c

15 Di-hadron Correlations STARLHC, ALICE acceptance HIJING Simulation “Peak Inversion” O (1)/2  events M. Ploskon, ALICE INT

16 The biased trigger bias hep-ph/ p T trig > 8 GeV is a function of p T trig but alsp p T assoc,  s, near-side/away-side,  E See also, K. Filimonov, J.Phys.G31:S513-S520,2005

17 From di-hadron correlations to jets Strong bias on fragmentation function … which we want to measure Low selectivity of the parton energy Very low efficiency, example: ~6% for E T > 100 GeV Jets produced in central Pb-Pb collisions (|  | < 0.5) No trigger: ~ Jets on tape ~1500 Jets selected using leading particles

18 Reduction of the trigger bias by collecting more energy from jet fragmentation… Unbiased parton energy fraction production spectrum induced bias

19 Reconstructed Jets: Objectives Reduce the trigger bias as much as possible by collecting of maximum of jet energy Maximum cone-radius allowed by background level Minimum p T allowed by background level Study jet structure inclusively Down to lowest possible p T (z, j T ) Collect maximum statistics using trigger.

20 Jet Finder in HI Environment:Principle Loop1: Background estimation from cells outside jet cones Loop2: UA1 cone algorithm to find centroid using cells after background subtraction RcRc

21 Jet Finder based on cone algorithms Input: List of cells in an  grid sorted in decreasing cell energy E i Estimate the average background energy E bg per cell from all cells. For at least 2 iterations and until the change in E bg between 2 successive iterations is smaller than a set threshold: Clear the jet list Flag cells outside a jet. Execute the jet-finding loop for each cell, starting with the highest cell energy. If E i – E bg > E seed and if the cell is not already flagged as being inside a jet: Set the jet-cone centroid to be the center of the jet seed cell (  c,  c ) = (  i,  i ) Using all cells with  (  i -  ) 2 +(  i -  ) 2 < R c of the initial centroid, calculate the new energy weighted centroid to be the new initial centroid. Repeat until difference between iterations shifts less than one cell. Store centroid as jet candidate. Recalculate background energy using information from cells outside jets.

22 Optimal Cone Size Jets reconstructed from charged particles: Need reduced cone sizes and transverse momentum cut ! Energy contained in sub-cone R E ~ R 2 Jet Finders for AA do not work with the standard cone size used for pp (R = 0.7-1). R and p T cut have to be optimized according to the background conditions.

23 Background Fluctuations Background fluctuations limit the energy resolution. Fluctuations caused by event-by-event variations of the impact parameter for a given centrality class. Strong correlation between different regions in  plane ~R 2 Can be eliminated using impact parameter dependent background subtrcation. Poissonian fluctuations of uncorrelated particles  E =  N  [ 2 +  p T 2 ] ~R Correlated particles from common source (low-E T jets) ~R Out-of-cone Fluctuations

24 Background Fluctuations Evt-by-evt background energy estimation

25 Signal fluctuations Response function for mono-chromatic jets E T = 100 GeV  E/E ~ 50%  E/E ~ 30%

26 Putting things together: Intrinsic resolution limit p T > 0 GeV 1 GeV 2 GeV Resolution limited by out-of-cone fluctuations common to all experiments ! E jet = 100 GeV Background included

27 Expected resolution including EMCAL Jet reconstruction using charged particles measured by TPC + ITS And neutral energy from EMCAL.

28 Trigger performance Trigger on energy in patch  x  Background rejection set to factor of 10 =>HLT Centrality dependent thresholds

29 Reference systems Jet trigger Compare central Pb+Pb to reference measurements Pb+Pb peripheral: vary system size and shape p+A: cold nuclear matter effects p+p (14 TeV): no nuclear effects, but different energy p+p (5.5 TeV): ideal reference, but limited statistics Includes acceptance, efficiency, dead time, energy resolution All reference systems are required for a complete systematic study

30 Jet yields: one LHC year Jet yield in 20 GeV bin Large gains due to jet trigger Large variation in statistical reach for different reference systems

31 Resolution buys statistics

32 ALICE performance What has been achieved so far ? Full detector simulation and reconstruction of HIJING events with embedded Pythia Jets Implementation of a core analysis frame work Reconstruction and analysis of charged jets. Quenching Studies on fragmentation function.

33 Energy spectrum from charged jets Cone-Algorithm: R = 0.4, p T > 2 GeV Selection efficiency ~30% as compared to 6% with leading particle ! No de-convolution, but Gauss  E -n ~ E -n

34 Jet structure observables Low z (high  ): Systematics is a challenge, needs reliable tracking. Also good statistics (trigger is needed)

35 Hump-back plateau Bias due to incomplete reconstruction. E rec > 100 GeV Statistical error 2 GeV 10 4 events

36 Systematics of background subtraction Background energy is systematically underestimated ( O (1 GeV)) Corrections under study (thesis work of R. Dias Valdez)

37 j T -Spectra Bias due to incomplete reconstruction. E rec > 100 GeV Statistical error 10 4 events  jTjT

38 Estimate quenching at LHC: Quenching Studies Compare distributions with and without quenching The measurement: ratio of dashed over solid = Pb+Pb(central)/p+p Solid: unquenched (p+p) Pythia-based simulation with quenching Large R, no p T cut Dashed: quenched jet (central Pb+Pb)

39 Toy Models Pythia hard scattering Initial and Final State Radiation Afterburner A Afterburner B Afterburner C Pythia Hadronization Two extreme approaches Quenching of the final jet system and radiation of 1-5 gluons. (AliPythia::Quench using Salgado/Wiedemann - Quenching weights) Quenching of all final state partons and radiation of many (~40) gluons (I. Lokhtin: Pyquen) * Nuclear Geometry (Glauber) ) * I.P. Lokhtin et al., Eur. Phys. J C16 (2000) I.P.Lokhtin et al., e-print hep-ph/ Jet (E) → Jet (E-  E) + n gluons (“Mini Jets”)

40 ALICE+EMCal in one LHC year ratio

41 Benchmark measurement: p+Pb reference With EMCal: jet trigger+ improved jet reconstruction provides much greater E T reach

42 Benchmark measurement: Peripheral Pb+Pb reference Without EMCal, significant quenching measurements beyond ~100 GeV are not possible

43 Summary of statistical reach Ratio  >4 With EMCALW/O EMCAL R AA R pA R AA (5.5 TeV) R AA (  ) R CP 150 (70) Ratio z>0.5With EMCALW/O EMCAL R AA R pA 150 (70) R AA (5.5 TeV)140 (60) Large  : ~10% error requires several hundred signal events (Pb central) and normalization events (pp,pA). Large z>0.5 requires several thousand events The EMCAL extends kinematic range by 40–125 GeV improves resolution (important at high z) Some measurements impossible w/o EMCAL

44 More to come … Dijet correlations “Sub-jet” Suppression ? Look for “hot spots” at large distance to jet axis ~10 GeV parton suppression within 100 GeV jets ? R 0 = 1fm  t form = 1/(  k T ) t sep = 1/ Q 

45 Photon-tagged jets Dominant processes: g + q → γ + q (Compton) q + q → γ + g (Annihilation) p T > 10 GeV/c   -jet correlation E  = E jet Opposite direction Direct photons are not perturbed by the medium Parton in-medium-modification through the fragmentation function  min  max IP PHOS EMCal TPC 

46 Identifying prompt  in ALICE x5 signal Statistics for on months of running: 2000  with E  > 20 GeV E  reach increases to 40 GeV with EMCAL

47 Fragmentation function quenched jet non-quenched Pb-Pb collisions Background Signal HIC background

48 Summary Copious production of jets in Pb-Pb collisions at the LHC < 20 GeV many overlapping jets/event Inclusive leading particle correlation Background conditions require jet identification and reconstruction in reduced cone R < At LHC we will measure jet structure observables (j T, fragmentation function, jet-shape) for reconstructed jets. High-p T capabilities (calorimetry) needed to reconstruct parton energy Good low-p T capabilities are needed to measure particles from medium induced radiation. EMCAL will provide trigger capabilities which are in particular needed to perform reference measurements (pA, pp,..) ALICE can measure photon tagged jets with E  > 20 GeV (PHOS + TPC) E  > 40 GeV (EMCAL+TPC) Sensitivity to medium modifications ~5%