measurement of jet quenching in pb-pb snn = …...2016/04/21 · hiroki yokoyama : 2nd year thesis...

Measurement of Jet Quenching in Pb-Pb Collisions at √sNN = 5.02 TeV in the ALICE Experiment at the LHCHiroki Yokoyama

LPSC, Université Grenoble-Alpes, CNRS/IN2P3 University of Tsukuba, Japan

21/04/2016 2nd year Thesis Monitoring Seminar

1

Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016

OutlineQuark-Gluon Plasma and Jet Quenching

Quark-Gluon PlasmaJets in Heavy Ion Collisions

The ALICE Experiment at the LHC Development and commissioning of the ALICE Calorimeter L1 Trigger system

L1-Jet Trigger Algorithmbackground subtraction methodTrigger Performance

Inclusive jet measurement with √sNN = 5.02 TeV Pb-Pb collisions Jet ReconstructionUnderlying EventUnfolding Technique

Summary

2

Quark-Gluon Plasma (QGP) and Jet Quenching

3


A brief journey to the beginning of the Universe

Quark-Gluon Plasma (QGP) Hot & dense color thermalized QCD matter prevailing at the early Universe ~1μs after big bangDeconfined state of quarks and gluonsTheoretically inferred through lattice gauge simulations of QCD

What is QGP ?

How to create ?‘Little Bang’

high-energy head-on nucleus-nucleus collisions at particle acceleratorsRecreate QGP droplets for a brief period of time to quantitatively map out the QCD phase diagram

4


Collimated spray of hadrons produced by the hard scattering of partons at the initial stage of the collisionhigh-Q2 process, pT ≳ 20 GeV

Jets in HI Collisions ( Hard Probes of the QGP )

Attenuation or disappearance of observed Jets in Pb-Pb

due to partons' energy loss in the QGPevaluation of the degree of the attenuation gives us QGP properties

What’s a Jet ? Jet Quenching

5

Why Jets ?The QGP lifetime is so short (~10-23s) that characterization by external probes is ruled out

self-produced probe Occur at early stage ~1/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)


ALICE Jet Quenching Measurements in Pb-PbNuclear modification factor : RAA

if RAA = 1, NO modification of the matter

)c (GeV/Tp

0 2 4 6 8 10 12 14 16 18 20

AAR

0.2

0.4

0.6

0.8

1=2.76 TeVNNsALICE 0-5% Pb-Pb

-π++π-+K+K

pp + Charged

ALI−DER−86149)c (GeV/

T,jetp

0 50 100

AAR

0

0.2

0.4

0.6

0.8

1

1.2 = 2.76 TeVNNsALICE Pb-Pb

Data 0 - 10%

JEWEL

YaJEM

Correlated uncertaintyShape uncertainty

)c (GeV/T,jetp50 100

| < 0.5jet

η = 0.2 | R TkAnti- c > 5 GeV/ lead,chTp

Data 10 - 30%

JEWEL

YaJEM


ALI−PUB−92182

high-pT hadrons strong suppression : RAA~0.2proxy for Jet (parton) : pT >10 GeV/cfragmentation pattern changed

Jets strong suppression : RAA~0.4Jet shape broadens? where is the lost energy?

6

R

AA

=1

< N

coll

>· #of particles(jets) in PbPb

#of particles(jets) in pp


ALICE Jet Quenching Measurements in Pb-PbNuclear modification factor : RAA

if RAA = 1, NO modification of the matter

)c (GeV/Tp

0 2 4 6 8 10 12 14 16 18 20

AAR

0.2

0.4

0.6

0.8

1=2.76 TeVNNsALICE 0-5% Pb-Pb

-π++π-+K+K

pp + Charged

ALI−DER−86149)c (GeV/

T,jetp

0 50 100

AAR

0

0.2

0.4

0.6

0.8

1

1.2 = 2.76 TeVNNsALICE Pb-Pb

Data 0 - 10%

JEWEL

YaJEM


)c (GeV/T,jetp50 100

| < 0.5jet

η = 0.2 | R TkAnti- c > 5 GeV/ lead,chTp

Data 10 - 30%

JEWEL

YaJEM


ALI−PUB−92182

high-pT hadrons strong suppression : RAA~0.2proxy for Jet (parton) : pT >10 GeV/cfragmentation pattern changed

Jets strong suppression : RAA~0.4Jet shape broadens? where is the lost energy?

7

R

AA

=1

< N

coll

>· #of particles(jets) in PbPb

#of particles(jets) in pp

Jet measurement in Pb-Pb collisions is challenging because of very large fluctuating background.

Need a dedicated trigger algorithm to select events containing jets [My first thesis topic]Extract a ‘clean signal’ despite pT smearing and “fake” jets [My second thesis topic]


Jet Measurement in LHC-ALICE

Calorimeters trigger Jet events &high-pT photon event

L1 trigger implementation

EMCal

DCal

PHOS

ALICE detector focus on Heavy Ion ExperimentLHC Run2 period started from 2015

√s = 13 TeV p-p√sNN = 5.02 TeV Pb-Pb

Charged Particles : |η| < 0.9, 0<φ<2π

EMCal, (DCal : Run2 from 2015-)Pb-Scintillator sampling calorimeter

PHOSlead-tungsten crystal (PWO) based calorimeter

Neutral constituents

Neutral particles : |η| < 0.7

ITS : silicon tracking detectorTPC : gas detector Charged constituents

Full Jet

8

PHOSCharged Jet

Development of the L1-Jet/photon Trigger System by Calorimeters

9


Search events with energy deposit larger than threshold in certain areaL1 Jet patch ( = 8x8 EMCal modules )

Amplitude calculation by sliding window algorithm

L1-Jet Trigger Algorithm

L1 jet patch = 8x8 modules

jet primitive = 4x4 modules

10

EMCal

η

φ

DCal DCalPHOS

7

3

11

4

0

8

15

12

2

27

35

32

Amplitudes of PHOS jet primitive are scaled complicated geometry of DCal+PHOS side


L1-Jet Background Subtraction Method in Pb-Pb

11

Large background energy deposit superimposed to the jet energyevent-by-event correction using the background measured on the opposite side

To avoid jet bias on background estimateTo obtain a reliable event-by-event background

EMCAL

Module BKG_density (EMCAL)

threshold

L1 out

8x8 Module Sum Calculate BKG

Subtract BKG Compare

L1 out

DCAL

Module BKG_density (DCAL)

threshold

8x8 Module Sum Calculate BKG

Subtract BKG Compare


PHOS TRUPHOS TRUPHOS TRUPHOS FEE

PHOS TRUPHOS TRU

L1 System of ALICE Calorimeters

PHOS STU

DCAL STU

CTPPHOS TRUPHOS TRU

DCAL TRUDCAL TRUDCAL TRUDCAL TRU

DCAL_L0 DCAL+PHOS_L1_Jet_low

DCAL+PHOS_L1_Jet DCAL_L1_photon_low

DCAL_L1_photon

L0_local fastOR amp

L0_local fastOR amp

subregion data*28TRUs

*14TRUs

DCAL TRUDCAL TRUDCAL TRUEMCAL

TRU

L0_local fastOR amp

*32TRUs

EMCAL STU

background density

EMCAL_L0 EMCAL_L1_Jet_low

EMCAL_L1_Jet EMCAL_L1_photon_low

EMCAL_L1_photon

PHOS_L0 PHOS_L1_photon_low

PHOS_L1_photon

DCAL TRUDCAL TRUDCAL TRUDCAL FEE

DCAL TRUDCAL TRUDCAL TRUEMCAL

FEE

fastOR amp (analog)

fastOR amp (analog)

fastOR amp (analog)

Analog sum inside 1fastOR

digitise fastOR Amp L0 trigger calculation

L1 photon/Jet Trigger calculation

Contribution of Grenoble Group :

Design / Production of the STU (Summary Trigger Unit)

FPGA FW development on STU at period of Run1 (2010-) Olivier. B

MY ACTIVITY :

FPGA FW development on STU for Run2 period (2015-)

Trigger system commissioning

12


Trigger PerformanceTrigger performance

Data produced by Pb-Pb run in 2015PHOS 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~1000 for L1-Jet @20GeV threshold

Threshold[GeV]40− 20− 0 20 40 60 80 100

Rej

ectio

n5−10

4−10

3−10

2−10

1−10

1 L1-Jet RejectionEMCalDCal

(EMCAL)[GeV]ρ0 50 100 150 200 250 300

(DCAL)[G

eV]

ρ

0

50

100

150

200

250

300

1

10

210

310

410

510

HIROKI YOKOYAMA

Jet Trigger Rejection

13

BKG density correlation

DCal

BKG

den

sity

EMCal BKG density

Trigger Threshold [GeV]


Trigger Performance

FEE amp in JP/GP [GeV]0 10 20 30 40 50 60 70 80

effic

ienc

y

0

0.2

0.4

0.6

0.8

1

1.2

HIROKI YOKOYAMA

EMCAL Jet_H

DCAL Jet_H

EMCAL Gamma_H

DCAL Gamma_H

Trigger efficiency clear turn-on at set values of 10 GeV (L1-photon) and 20 GeV (L1_Jet)

All L1-Jet/photon Triggers by Calorimeters Worked Well in 2015 Pb-Pb Runs!

14

Data produced by Pb-Pb run in 2015PHOS 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~1000 for L1-Jet @20GeV threshold

effici

ency

Energy in Jet/photon patch

Measurement of the Inclusive Jet with √sNN = 5.02 TeV Pb-Pb Collisions

15


(1) Jet ReconstructionCorrespondence between detector level, hadron level and parton level.Jet Reconstruction

using observable particlessequentially combine or classify these particles into clusters (Jet candidates)

Some types of algorithms

16


Analysis Strategies

Charged Particle Selection

Jet Reconstruction

Underlying Event background subtraction

Unfolding

Charged Particle Selection

Neutral Particle Selection

Jet Reconstruction

Unfolding

Inclusive Jet spectra Inclusive Jet spectra

MB Trigger event, Charged Jet MB, Jet Trigger event, Full Jet (Charged+Neutral)

Data : Pb-Pb (√sNN = 5.02 TeV) taken in 2015 Nov. and Dec.

17

Underlying Event background subtraction(2)

(3)

(1)


Estimated BKG is subtracted from each Jetevent-by-event basis via

background density : ρmedian of kT clusters except largest two:

large background superimposed to the JetsdNch/dη ~ 2000 in most central collisions

[GeV/c]jetT

p0 50 100 150 200 250

T/d

pje

t d

Nev

1/N

5−10

4−10

3−10

2−10

1−10

1

10

210=5.02TeV, cent: 0-10%NNsPbPb

before BKG subtraction

after BKG subtraction

10

210

310

Centrality (%)0 10 20 30 40 50 60 70 80 90

(GeV

/c)

ρ

0

50

100

150

200

250

my result

Difficulties in Heavy Ion Collision

Jet Background Subtraction ← charged jet BKG density vs. centrality

↓ Jet pT spectra before/after BKG subtraction (R=0.4)

my result

18

(2) Underlying Event

0-10 % Central eventsbefore subtractionafter subtraction

~50 GeV BKG in the Jet Area


Background Fluctuations Detector Effects

(3) UnfoldingMore correction NEEDED!

Detector effect (e.g. smearing) background fluctuation

Unfolding technique

Mm = Gm,d ·Dd Dd = Gd,t · Tt

Measuredjet spectrum

Responsematrix

“unknown” truejet spectrum

Responsematrix

spectrum corrected for BKG fluctuation

Inverted Response matrix gives “TRUE” jet spectrum via

Mm = Gm,d ·Gd,t · Tt = Gm,t · Tt

19

MC simulation


SummaryQuark-Gluon Plasma (QGP)

Deconfined state of quarks and gluonsCreated in “Little Bang” ( Heavy Ion Collisions )

Jet quenching Attenuation of the energy of hard scattered partonBest equivalent to quarks/gluons

Development of L1-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. 2016Full Jet Analysis : depend on the EMCal calibration status

20

BACKUP

21


Relativistic Heavy Ion Collision

Hard parton scatteringhigh momentum parton production

Creation of a Quark-Gluon Plasmathermalisation of strongly interacting partonscollective / thermodynamical properties

Phase transitionfrom parton phase to hadron phaseTc ≈ 155MeV, εc ≈ 0.5GeV/fm3

Hadronic phasefix species, number and momentum of hadrons

Final particles (hadrons, leptons) are detected by particle detectorC.)NONAKA)) ®½Ä°¾�¿KO�)

´±ÁÂ�ÃUC-H��F�thermaliza/on� hydro� hadroniza/on� freezeout�collisions�

22


Modelling of the Jet Quenching

Radiative energy loss (gluon bremsstrahlung) is dominantRelative strength is controlled by Casimir factor CR

αsCF : q→qgαsCA : g→ggαsTF : g→qqbar

ΔE measurement provides information on matter properties (Jet tomography)

23


Jet / charged particle spectra in Pb-Pb and p-p collisions

(GeV/c)T

p0 5 10 15 20

-2) (

GeV

/c)

T d

pη

) / (d

chN2) (

dT

pπ

1/(2

evt

1/N

-810

-710

-610

-510

-410

-310

-210

-110

1

10

210

310

410

510

scaled pp reference0-5%70-80%

= 2.76 TeVNNsPb-Pb

ALI−PUB−86029

AWAITI

NG APP

ROVAL

)c (GeV/T,jetp0 50 100

-1 )c (G

eV/

jet

ηdT,

jet

pdN2 d

ev

tN1

coll

N1

-910

-810

-710

-610

= 2.76 TeVNNsALICE

| < 0.5jet

η = 0.2 | R TkAnti-c > 5 GeV/lead,ch

Tp


pp0 - 10% Pb-Pb10 - 30% Pb-Pb

ALI−PUB−92515

24


FEE, TRU and STUFEE (Front-End Electronics card)

mainly, calculate analog sum of 2x2 towers amplitudefastOR ( module ) = 2x2 towers area

TRU ( Trigger Region Unit )digitise analog fastOR amplitudeL0 calculation

STU ( Summary Trigger Unit )L1-Jet/Photon calculation

FEE TRU STU

25


L1-photon patch 2x2 [modules] (EMCAL, DCAL) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm

L1 photon/Jet algorithm in Run20 1 2 3 4 5 6 7

8 9 10 11 12 13 14 15

16 17 18 19 20 21 22 23

24 25 26 27 28 29 30 31

32 33 34 35 36 37 38 39

40 41 42 43 44 45 46 47

48 49 50 51 52 53 54 55

56 57 58 59 60 61 62 63

64 65 66 67 68 69 70 71

72 73 74 75 76 77 78 79

80 81 82 83 84 85 86 87

88 89 90 91 92 93 94 95

0 1 2 3 4 5 6 7

8 9 10 11 12 13 14 15

16 17 18 19 20 21 22 23

24 25 26 27 28 29 30 31

32 33 34 35 36 37 38 39

40 41 42 43 44 45 46 47

48 49 50 51 52 53 54 55

56 57 58 59 60 61 62 63

64 65 66 67 68 69 70 71

72 73 74 75 76 77 78 79

80 81 82 83 84 85 86 87

88 89 90 91 92 93 94 95

0 1 2 3 4 5 6 7

8 9 10 11 12 13 14 15

16 17 18 19 20 21 22 23

24 25 26 27 28 29 30 31

32 33 34 35 36 37 38 39

40 41 42 43 44 45 46 47

48 49 50 51 52 53 54 55

56 57 58 59 60 61 62 63

64 65 66 67 68 69 70 71

72 73 74 75 76 77 78 79

80 81 82 83 84 85 86 87

88 89 90 91 92 93 94 95

0 1 2 3 4 5 6 7

8 9 10 11 12 13 14 15

16 17 18 19 20 21 22 23

24 25 26 27 28 29 30 31

32 33 34 35 36 37 38 39

40 41 42 43 44 45 46 47

48 49 50 51 52 53 54 55

56 57 58 59 60 61 62 63

64 65 66 67 68 69 70 71

72 73 74 75 76 77 78 79

80 81 82 83 84 85 86 87

88 89 90 91 92 93 94 95

26


L1 photon/Jet algorithm in Run2L1-photon patch

2x2 [modules] (EMCAL, DCAL) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm

L1-Jet patch (jet-primitive = 4x4 modules) 2x2 jet-primitives sliding window algorithm

27


L1 photon/Jet algorithm in Run2L1-photon patch

2x2 [modules] (EMCAL, DCAL) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm

L1-Jet patch (jet-primitive = 4x4 modules) 2x2 jet-primitives sliding window algorithm

L1-Jet background BKG patch = 2x2 jet-primitives no overlap BKG = median of BKG patch amp

Epatch - median{EBKG/ABKG} * Apatch > Threshold => L1 activated

28


EMCAL

DCAL

L1-Jet Background Subtraction Method

BKG Jet patch (no overlapped 2x2 jet

primitive)jet primitive calc

BKG density calculation

(median calc)

Module amplitude BKG_density(EMCAL)

BKG subtraction

BKG Jet patch (no overlapped 2x2 jet

primitive)jet primitive calc

BKG density calculation

(median calc)

Module amplitude BKG_density(DCAL)

BKG subtraction

threshold

comparison L1 out

threshold

comparison L1 out

jet primitive

BKG jet patch

BKG jet patch

Jet patch calc (nxn jet primitive)

Jet patch calc (nxn jet primitive)

jet patch

jet patch

PHOS_subregion

jet primitive

jet primitive

jet primitive

29


Trigger Rejection, Trigger EfficiencyTrigger performance

Threshold[GeV]40− 20− 0 20 40 60 80 100

Rej

ectio

n

5−10

4−10

3−10

2−10

1−10

1 L1-Jet RejectionEMCalDCal

Threshold[GeV]10− 5− 0 5 10 15 20 25 30

Rej

ectio

n

5−10

4−10

3−10

2−10

1−10

1 L1-Photon Rejection

EMCal

DCal

Photon Trigger Jet Trigger

FEE amp in JP/GP [GeV]0 10 20 30 40 50 60 70 80

eff

icie

ncy

0

0.2

0.4

0.6

0.8

1

1.2

HIROKI YOKOYAMA

EMCAL Jet_H

DCAL Jet_H

EMCAL Gamma_H

DCAL Gamma_H

Trigger E�ciency =Number of Jet Patch in Triggered Events

Number of Jet Patch in MB Events

30


Jet Reconstruction AlgorithmFastJet anti-kT algorithm ( p=-1, p=1 for kT algorithm)

calculate dij and diB by all particles combinationwhen 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 jetrepeat until no particle are left

31


Definition of the Underlying Energy Density1. 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 listbecause (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

32


Underlying Event @ other centralityat the peripheral collision, background subtraction don’t have impact due to the small background

[GeV/c]jetT

p0 50 100 150 200 250

T/d

pje

t d

Nev

1/N

5−10

4−10

3−10

2−10

1−10

1

10




[GeV/c]jetT

p0 50 100 150 200 250

T/d

pje

t d

Nev

1/N

5−10

4−10

3−10

2−10

1−10

1

10




33

measurement of jet quenching in pb-pb snn = …...2016/04/21 · hiroki yokoyama : 2nd year thesis...

Documents