Measurement of Jet Quenching in Pb-Pb Collisions at snn = 5.02 TeV in the ALICE Experiment at the LHC

Transcription

1 Measurement of Jet Quenching in Pb-Pb Collisions at snn = 5.02 TeV in the ALICE Experiment at the LHC Hiroki Yokoyama LPSC, Université Grenoble-Alpes, CNRS/IN2P3 University of Tsukuba, Japan 2/04/206 2nd year Thesis Monitoring Seminar

2 2 Outline Quark-Gluon Plasma and Jet Quenching Quark-Gluon Plasma Jets in Heavy Ion Collisions The ALICE Experiment at the LHC Development and commissioning of the ALICE Calorimeter L Trigger system L-Jet Trigger Algorithm background subtraction method Trigger Performance Inclusive jet measurement with snn = 5.02 TeV Pb-Pb collisions Jet Reconstruction Underlying Event Unfolding Technique Summary

3 Quark-Gluon Plasma (QGP) and Jet Quenching 3

4 4 A brief journey to the beginning of the Universe What is QGP? Quark-Gluon Plasma (QGP) Hot & dense color thermalized QCD matter prevailing at the early Universe ~μs after big bang Deconfined state of quarks and gluons Theoretically inferred through lattice gauge simulations of QCD Little Bang How to create? high-energy head-on nucleus-nucleus collisions at particle accelerators Recreate QGP droplets for a brief period of time to quantitatively map out the QCD phase diagram

5 5 Jets in HI Collisions ( Hard Probes of the QGP ) What s a Jet? Collimated spray of hadrons produced by the hard scattering of partons at the initial stage of the collision high-q 2 process, pt 20 GeV Why Jets? The QGP lifetime is so short (~ -23 s) that characterization by external probes is ruled out self-produced probe Occur at early stage ~/Q probe the entire medium evolution Production rate calculable within pqcd well calibrated probe Large cross-section at the LHC copious production Reconstructed Jet enables to access 4-momentum of original parton jet structure (energy re-distribution) Jet Quenching Attenuation or disappearance of observed Jets in Pb-Pb due to partons' energy loss in the QGP evaluation of the degree of the attenuation gives us QGP properties

6 6 ALICE Jet Quenching Measurements in Pb-Pb Nuclear modification factor : RAA if RAA =, NO modification of the matter R AA = < N coll > #of particles(jets) in PbPb #of particles(jets) in pp high-pt hadrons strong suppression : RAA~0.2 proxy for Jet (parton) : pt > GeV/c fragmentation pattern changed Jets strong suppression : RAA~0.4 Jet shape broadens? where is the lost energy? R AA ALICE 0-5% Pb-Pb - π + +π + - K +K p + p Charged s NN =2.76 TeV R AA ALICE Pb-Pb Data 0 - % JEWEL YaJEM s NN = 2.76 TeV Correlated uncertainty Shape uncertainty Anti-k T R = 0.2 η < 0.5 jet Data - 30% JEWEL YaJEM lead,ch p T > 5 GeV/ c Correlated uncertainty Shape uncertainty ALI DER (GeV/c) p T p (GeV/c) T,jet ALI PUB (GeV/ c) p T,jet

7 7 ALICE Jet Quenching Measurements in Pb-Pb Nuclear modification factor : RAA if RAA =, NO modification of the matter R AA = < N coll > #of particles(jets) in PbPb #of particles(jets) in pp R AA high-pt hadrons proxy for Jet (parton) : pt > GeV/c.2 ALICE Pb-Pb s NN = 2.76 TeV Anti- lead,ch Extract a clean signal despite pt smearing and fake k T jets R = 0.2 η < 0.5 p > 5 GeV/ c jet T [My second ALICE thesis 0-5% Pb-Pb topic] s NN =2.76 TeV Correlated uncertainty Correlated uncertainty - π + Data 0 - % Data - 30% +π Shape uncertainty Shape uncertainty K +K JEWEL JEWEL p + p YaJEM YaJEM Charged 0.6 R AA Jets Jet measurement in Pb-Pb collisions is challenging because of very large fluctuating strong suppression background. : RAA~0.2 strong suppression : RAA~0.4 Jet shape broadens? where is the lost energy? Need fragmentation a dedicated pattern trigger changed algorithm to select events containing jets [My first thesis topic] ALI DER (GeV/c) p T p (GeV/c) T,jet ALI PUB (GeV/ c) p T,jet

8 8 Jet Measurement in LHC-ALICE ALICE detector focus on Heavy Ion Experiment LHC Run2 period started from 205 s = 3 TeV p-p s NN = 5.02 TeV Pb-Pb Charged Particles : η < 0.9, 0<φ<2π ITS : silicon tracking detector TPC : gas detector Charged constituents PHOS Full Jet Neutral particles : η < 0.7 EMCal Charged Jet EMCal, (DCal : Run2 from 205-) PHOS Pb-Scintillator sampling calorimeter lead-tungsten crystal (PWO) based calorimeter Neutral constituents PHOS DCal Calorimeters trigger Jet events & high-pt photon event L trigger implementation

9 Development of the L-Jet/photon Trigger System by Calorimeters 9

10 L-Jet Trigger Algorithm Search events with energy deposit larger than threshold in certain area L Jet patch ( = 8x8 EMCal modules ) Amplitude calculation by sliding window algorithm EMCal jet primitive = 4x4 modules L jet patch = 8x8 modules Amplitudes of PHOS jet primitive are scaled complicated geometry of DCal+PHOS side DCal PHOS DCal η φ

11 L-Jet Background Subtraction Method in Pb-Pb Large background energy deposit superimposed to the jet energy event-by-event correction using the background measured on the opposite side To avoid jet bias on background estimate To obtain a reliable event-by-event background EMCAL Module 8x8 Module Sum Calculate BKG BKG_density (EMCAL) Subtract BKG Compare L out threshold Module 8x8 Module Sum Calculate BKG BKG_density () Subtract BKG Compare L out threshold

12 2 L System of ALICE Calorimeters PHOS PHOS PHOS FEE TRU PHOS TRU TRU fastor amp (analog) PHOS PHOS PHOS TRU TRU PHOS TRU TRU L0_local fastor amp PHOS STU PHOS_L0 PHOS_L_photon_low PHOS_L_photon CTP *28TRUs subregion data FEETRU TRU TRU fastor amp (analog) TRU TRU TRU TRU L0_local fastor amp STU _L0 +PHOS_L_Jet_low +PHOS_L_Jet _L_photon_low _L_photon Contribution of Grenoble Group : Design / Production of the STU (Summary Trigger Unit) *4TRUs background density FPGA FW development on STU at period of Run (20-) Olivier. B EMCAL FEETRU TRU TRU fastor amp (analog) EMCAL TRU TRU TRU TRU L0_local fastor amp EMCAL STU EMCAL_L0 EMCAL_L_Jet_low EMCAL_L_Jet EMCAL_L_photon_low EMCAL_L_photon MY ACTIVITY : FPGA FW development on STU for Run2 period (205-) Analog sum inside fastor *32TRUs digitise fastor Amp L0 trigger calculation L photon/jet Trigger calculation Trigger system commissioning

13 3 Trigger Performance Data produced by Pb-Pb run in 205 PHOS didn t join in whole Jet Trigger calculation due to insufficient tune of TRUs BKG density correlation between EMCal and DCal Reasonable positive correlation Trigger rejection To define threshold which satisfy the BandWidth restriction of the data taking ~00 for threshold DCal BKG density ρ()[gev] BKG density correlation HIROKI YOKOYAMA ρ(emcal)[gev] EMCal BKG density Jet Trigger Rejection Rejection L-Jet Rejection EMCal DCal Threshold[GeV] Trigger Threshold [GeV]

14 4 Trigger Performance Data produced by Pb-Pb run in 205 PHOS didn t join in whole Jet Trigger calculation due to insufficient tune of TRUs BKG density correlation between EMCal and DCal Reasonable positive correlation Trigger rejection To define threshold which satisfy the BandWidth restriction of the data taking ~00 for threshold Trigger efficiency clear turn-on at set values of GeV (L-photon) and 20 GeV (L_Jet) efficiency efficiency EMCAL Jet_H Jet_H EMCAL Gamma_H Gamma_H HIROKI YOKOYAMA FEE amp in JP/GP [GeV] Energy in Jet/photon patch All L-Jet/photon Triggers by Calorimeters Worked Well in 205 Pb-Pb Runs!

15 Measurement of the Inclusive Jet with snn = 5.02 TeV Pb-Pb Collisions 5

16 6 () Jet Reconstruction Correspondence between detector level, hadron level and parton level. Jet Reconstruction using observable particles sequentially combine or classify these particles into clusters (Jet candidates) Some types of algorithms

17 7 Analysis Strategies Data : Pb-Pb ( snn = 5.02 TeV) taken in 205 Nov. and Dec. MB Trigger event, Charged Jet Charged Particle Selection MB, Jet Trigger event, Full Jet (Charged+Neutral) Charged Particle Selection Neutral Particle Selection Jet Reconstruction () Jet Reconstruction Underlying Event background subtraction (2) Underlying Event background subtraction Unfolding (3) Unfolding Inclusive Jet spectra Inclusive Jet spectra

18 8 (2) Underlying Event Difficulties in Heavy Ion Collision large background superimposed to the Jets dnch/dη ~ 2000 in most central collisions ρ (GeV/c) Jet Background Subtraction background density : ρ median of kt clusters except largest two: 50 0 my result Centrality (%) 2 charged jet BKG density vs. centrality Jet pt spectra before/after BKG subtraction (R=0.4) /dp jet /N ev dn T 2 0- % Central events PbPb s NN =5.02TeV, cent: 0-% before subtraction BKG after subtraction BKG Estimated BKG is subtracted from each Jet event-by-event basis via 2 3 ~50 GeV BKG in the Jet Area 4 my result jet p [GeV/c] 250 T

19 9 (3) Unfolding More correction NEEDED! Detector effect (e.g. smearing) background fluctuation MC simulation Unfolding technique Background Fluctuations M m = G m,d D d Detector Effects D d = G d,t T t Measured jet spectrum Response matrix spectrum corrected for BKG fluctuation Response matrix unknown true jet spectrum Inverted Response matrix gives TRUE jet spectrum via M m = G m,d G d,t T t = G m,t T t

20 20 Summary Quark-Gluon Plasma (QGP) Deconfined state of quarks and gluons Created in Little Bang ( Heavy Ion Collisions ) Jet quenching Attenuation of the energy of hard scattered parton Best equivalent to quarks/gluons Development of L-Jet/photon trigger system by ALICE Calorimeters Background subtraction method was implemented on STU. All work was achieved. Measurement of Inclusive Jets in HI collisions Need some experimental technique ( jet reconstruction, underlying event subtraction and unfolding ) work in progress Charged Jet Spectra and nuclear modification factor : until Sep. 206 Full Jet Analysis : depend on the EMCal calibration status

21 BACKUP 2

22 22 Relativistic Heavy Ion Collision collisions thermaliza/on hydro hadroniza/on freezeout Hard parton scattering high momentum parton production Creation of a Quark-Gluon Plasma thermalisation of strongly interacting partons collective / thermodynamical properties Phase transition from parton phase to hadron phase T c 55MeV, ε c 0.5GeV/fm 3 Hadronic phase fix species, number and momentum of hadrons Final particles (hadrons, leptons) are detected by particle detector

23 23 Modelling of the Jet Quenching Radiative energy loss (gluon bremsstrahlung) is dominant Relative strength is controlled by Casimir factor CR αscf : q qg αsca : g gg αstf : g qqbar ΔE measurement provides information on matter properties (Jet tomography)

24 24 Jet / charged particle spectra in Pb-Pb and p-p collisions -2 ) (GeV/c) N ch ) / (dη dp 2 ) (d T T p /(2π /N evt ALI PUB Pb-Pb scaled pp reference 0-5% 70-80% = 2.76 TeV (GeV/c) s NN AWAITING APPROVAL p T - c) (GeV/ N 2 d jet dη T,jet dp N evt N coll ALI PUB 9255 ALICE s NN = 2.76 TeV Anti-k T lead,ch p T pp 0 - % Pb-Pb - 30% Pb-Pb R = 0.2 η < 0.5 jet > 5 GeV/ c Correlated uncertainty Shape uncertainty p T,jet (GeV/ c)

25 25 FEE, TRU and STU FEE (Front-End Electronics card) mainly, calculate analog sum of 2x2 towers amplitude fastor ( module ) = 2x2 towers area TRU ( Trigger Region Unit ) digitise analog fastor amplitude L0 calculation STU ( Summary Trigger Unit ) L-Jet/Photon calculation FEE TRU STU

26 26 L photon/jet algorithm in Run2 L-photon patch 2x2 [modules] (EMCAL, ) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm

27 27 L photon/jet algorithm in Run2 L-photon patch 2x2 [modules] (EMCAL, ) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm L-Jet patch (jet-primitive = 4x4 modules) 2x2 jet-primitives sliding window algorithm

28 28 L photon/jet algorithm in Run2 L-photon patch 2x2 [modules] (EMCAL, ) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm L-Jet patch (jet-primitive = 4x4 modules) 2x2 jet-primitives sliding window algorithm L-Jet background BKG patch = 2x2 jet-primitives no overlap BKG = median of BKG patch amp Epatch - median{ebkg/abkg} * Apatch > Threshold => L activated

29 29 L-Jet Background Subtraction Method Module amplitude jet primitive calc jet primitive BKG Jet patch (no overlapped 2x2 jet primitive) BKG jet patch BKG density calculation (median calc) EMCAL BKG_density(EMCAL) threshold jet primitive Jet patch calc (nxn jet primitive) jet patch BKG subtraction comparison L out Module amplitude jet primitive calc jet primitive BKG Jet patch (no overlapped 2x2 jet primitive) BKG jet patch BKG density calculation (median calc) BKG_density() threshold jet primitive Jet patch calc (nxn jet primitive) jet patch BKG subtraction comparison L out PHOS_subregion

30 30 Trigger Rejection, Trigger Efficiency Trigger E ciency = Number of Jet Patch in Triggered Events Number of Jet Patch in MB Events Photon Trigger Jet Trigger Rejection L-Photon Rejection EMCal DCal Rejection L-Jet Rejection EMCal DCal efficiency.2 EMCAL Jet_H Jet_H EMCAL Gamma_H 0.8 Gamma_H Threshold[GeV] Threshold[GeV] HIROKI YOKOYAMA FEE amp in JP/GP [GeV]

31 3 Jet Reconstruction Algorithm FastJet anti-kt algorithm ( p=-, p= for kt algorithm) calculate dij and dib by all particles combination when minimum d among them is part of dij merge particle i and j when minimum d among them is part of dib that cluster defined as jet repeat until no particle are left

32 32 Definition of the Underlying Energy Density. Output of kt algorithm : the list of the kt clusters ( pi, Ai ) kt algorithm is better than anti-kt algorithm for background estimation because low-pt particles are apt to be the cluster constituents. 2. Remove largest two clusters from the list because (2nd) largest cluster is the real Jet candidate. 3. Calculate median of pti / Ai median is better than average because the average calculation tend to be affected by seeped real-jet signal

33 33 Underlying other centrality at the peripheral collision, background subtraction don t have impact due to the small background /dp T 2 PbPb s NN =5.02TeV, cent: 0-% /dp T 2 PbPb s NN =5.02TeV, cent: 70-90% jet dn before BKG subtraction jet dn before BKG subtraction /N ev after BKG subtraction /N ev after BKG subtraction jet p [GeV/c] 250 T jet p [GeV/c] 250 T