yen-jie lee (cern) for the cms collaboration rencontres ion lourds/heavy ion meeting

1Study of jet quenching using the CMS detectorYen-Jie Lee (CERN)

Yen-Jie Lee (CERN)

for the CMS Collaboration

Rencontres Ion Lourds/Heavy Ion Meeting

18 April, 2013

Study of Jet Quenching using the CMS Detector


Probe the Medium

Final goal: Understand the

thermodynamics and transport

properties of QGP

Problem: the lifetime of QGP is so short

(O(fm/c)) such that it is not yet feasible

to probe it with an external source

Solution: Take the advantage of the

large cross-sections of hard probes

produced with the collision

QGP

source


Probe the Medium



properties of QGP







Colorless probes: γ, W/Z bosons

γ/W/Z

QGP


Probe the Medium



properties of QGP







Colorless probes: γ, W/Z bosons

Jets: originating from quarks and gluons

γ/W/Z

Jet

Jet

QGP

QGP

In medium parton energy loss

“Jet quenching”(Bjorken, 1982)


CMS Detector

Inner tracker:charged particlesvertex, isolation

solenoid

EM and Hadron calorimetersphotons, isolation

MuonMuon

HCALHCAL

ECALECAL

TrackerTracker |η|< 2.5

|η|< 3.0

|η|< 5.2

|η|< 2.4

Pb

Pb

Calojet

Particle Flow Jet (track pT> 0.9GeV/c)


Heavy Ion Collision Recorded by the CMS Detector

2010 PbPb 7 μb-1

2011 PbPb 150 μb-1

2012 pPb 1 μb-1

2013 pPb 31 nb-1


“Jet Quenching” without jet: Charged Particle RAA

Provide constraints on the parton energy loss models

R AA=σpp

inel

⟨N coll⟩

d2N AA / dpT dη

d2σ pp/ dpT dη

“QC

D M

ediu

m”

“QC

D V

acuu

m”

~

Ncoll

validate by photons

W/Z bosons

If PbPb = superposition of pp ...

EPJC 72 (2012) 1945


Charged Particle Spectra

Absorption? Energy loss?

Single hadron spectra themselves do not provide details of the underlying mechanism

Need direct jet reconstruction and correlation studies


Direct jet reconstruction with CMS

Subleading Jetp

T2

Leading Jetp

T1

Dijet Event Recorded by CMS


To Explain the Suppression of High pT Particles

Do we see strong suppression of high pT jets?Can we collect the radiated energy back?

Soft collinear radiation Hard radiation

PYTHIA inspired models Modified splitting functions

GLV + others(pre-LHC models)

Large angle soft radiation“QGP heating”

AdS/CFTInterference


Inclusive Jet Spectra: Jet RAA

Detector effects unfoldedStrong suppression of inclusive high pT jets

Compare PbPb to pp data

0.5

Anti-kT jets with R = 0.3

If PbPb = superposition of pp

CMS PAS HIN-12-004


Inclusive Jet Spectra: Jet RAA

Strong suppression of inclusive high pT jetsA cone of R=0.2, 0.3, 0.4 doesn’t catch all the radiated energyAre those high pT jets “completely absorbed” by the medium?

Compare PbPb to pp data

0.5

Anti-kT jets with R = 0.2, 0.3, 0.4

If PbPb = superposition of pp

CMS PAS HIN-12-004


Dijet and Photon-Jet Energy Imbalance

High pT leading jettriggered sample

High pT photontriggered sample

Photon unmodified jet energy tag

High statistics, with surface bias Lower statistics, without surface bias

Dijet Photon-jet


Dijet Momentum Imbalance

Parton energy loss is observed as a pronounced energy

imbalance in central PbPb collisions

Small AJ

(Balanced dijet)

Large AJ

(Un-balanced dijet)

pp

Jet Cone size R = 0.5

PRC 84 (2011) 024906


Dijet Energy Ratio (imbalance)

PLB 712 (2012) 176

Anti-kT jet R = 0.3

Dijet pT ratio as a function of leading jet pT


Dijet Energy Ratio (imbalance)

• Energy imbalance increases with centrality• Very high pT jets are also quenched PLB 712 (2012) 176

Anti-kT jet R = 0.3


-jet Momentum Imbalance

• Photons serve as an unmodified energy tag for the jet partner• Ratio of the pT of jets to photons (xJ=pT

Jet/pT)

is a direct measure of the jet energy loss• Gradual centrality-dependence of the xJ distribution

PLB 718 (2013) 773

PbPbPbPb

PbPbPbPb

Anti-kT jet R = 0.3


Dijet and photon-jet azimuthal correlation

Given the large momentum imbalanceseen in dijet and photon-jet events

Is the azimuthal correlation modified?

ΔφJet

Jet

QGP

In medium parton energy loss

“Jet quenching”


Dijet Azimuthal Angle Correlations

No apparent modification in the dijet Δφ distribution

for different jet pT (still back-to-back)

Δφ

Jet Cone size R = 0.5PLB 712 (2012) 176

0-20%


Photon-jet Azimuthal Angular Correlation

pp

Azimuthal angle difference between photon and jet

PbPb

pppp

“QGP Rutherfold experiment”

Jet

Jet

Photon

Photon“Backscattering?”

The first photon-jet correlation measurement in heavy ion collisions

PLB 718 (2013) 773

Anti-kT jet R = 0.3


Jet Shape and Fragmentation Function

Differential jet shape: “shape” of the jet as a function of radius (r)

Jet fragmentation function: how transverse momentum is distributed inside the jet cone

Tracks

Large parton energy loss (O(10GeV)) in the medium, out of the jet cone

What about jet structure? Measured using tracks in the jet cone.

r = (Δη2+Δφ2)1/2


PbPbPbPb PbPbPbPb

Differential Jet Shape

Significant modification at large radius (r) with respect to the jet axis, looking at tracks with pT> 1 GeV/c

r = (Δη2+Δφ2)1/2CMS PAS HIN-12-013


Jet Fragmentation Functions

PbPbPbPbPbPbPbPb

Low pT

particlesHigh pT

particlesInside the jet cone: Enhancement of low pT particle

Suppression of intermediate pT particles in cone

CMS PAS HIN-12-013


Track pT Distributions in Jet Cones (R=0.3)

High pT : no change compared to jets in pp collisionsIn (central) PbPb: excess of 1-2 tracks compared to pp at low pT

CMS PAS HIN-12-013

PbPbPbPbPbPbPbPb

(1/G

eV

)


What have we learned so far? (1)

3. Angular correlation of jets not largely modified

2. Large average dijet andphoton-jet pT imbalance

4. pT difference found at low pT particles far away

from the jets

5. Observation of modified jet fragmentation function and jet shape

Inside the jet cone:excess of ~1-2 tracks in PbPb

compared to pp at track pT < 3 GeV

What Have We Learned with CMS PbPb Data?

1. High pT jet suppressionΔR = 0.2 - 0.4 doesn’t capture all the radiated energy

6. b jets are also quenched(not included in this talk)


The Emerging Picture

pp PbPb

Excess of low pT particles, extends to ΔR>0.8

The bulk is not largely modified

Jet Energy loss ~5-15% in 0-20% central collisions

Excess of ~1-2 charged particles with pT < 4 GeV/cInside the jet cone R<0.3


Coming Back to the Three Scenarios

Soft collinear radiation Hard radiation

PYTHIA inspired models Modified splitting functions

GLV + others(pre-LHC models)

Large angle soft radiation“QGP heating”

AdS/CFTInterference


Summary Before reaching the final goal: understand the properties of QGP, we

are in the position to validate the theoretical understanding of the in-medium parton energy loss

CMS has presented interesting results from dijet, photon-jet and inclusive jet analyses in heavy ion collisions. A detailed picture of jet quenching is emerging.

To go beyond qualitative observation: An iterative feedback cycle between theory (in the form of MC

generator) and experiment is very important

To compare between data and theory: A proper smearing procedure for theorist (a proper unfolding

procedure for experimentalist when applicable) is needed

Summary


Backup slides


Summary Jet reconstruction and background subtraction: Improve jet reconstruction to account for elliptic flow using

forward calorimeter Analysis plan: Finalize QM12’ results

pPb data: Jet quenching in pPb collisions? Shadowing effects in pPb collisions? Corrected inclusive jet spectra in pp, pPb and PbPb

collisions PbPb data:

Further studies on dijet and photon-jet events and compare with high statistics pp sample (~5/pb) collected in 2013

Flavour dependence of jet quenching: study of multijet production and b-jet

Longer term: W/Z+jet analysis

Plan


Systematic uncertainties considered in analysis

X negligible/small effect, * important systematics, ** dominant systematics


How do we extract the medium effect in PbPb collisions?

RAA < 1 (suppression)

“QCD Medium”

“QCD Vacuum”

RAA > 1 (enhancement)

RAA = 1 (no medium effect) ~

pp reference PbPb measurements

One typical way is to compare PbPb data to pp reference measurement

‘Nuclear modification factors’

<Ncoll> Averaged number of binary scattering

R AA=σpp

inel

⟨N coll⟩

d2N AA / dpT dη

d2σ pp/ dpT dη


Background Subtraction

η

φ

-2.0 2.0-ππ

Exc

lude

η reflection MethodJetBkg

η

φ

-2.0 2.0-ππ

Event Mixing Method

Jet Bkg

η

φ

-2.0 2.0-ππ

Jet Event

Jet Event MinBias Event

Main result

(Cross-checks)


Tagging and Counting b-quark Jets Test the theoretical prediction color factor and quark-mass

dependence of in-medium parton energy loss

Secondary vertex tagged using

flight distance significance

Tagging efficiency estimated

in a data-driven way

Purity from template fits

to (tagged) secondary vtx

mass distributions

CMS PAS HIN-12-003


Pb+Pbp+p

Fraction of b-jets among All Jets

b-jet fraction: similar in pp and PbPb → b-jet quenching is comparable to light-jet quenching (RAA0.5), within present systematics

CMS PAS HIN-12-003


Jet Reconstruction and Composition

Anti-kT algorithm is used in most CMS publication

On average, charged hadrons carry 65% of the jet momentum

Measure the known part Correct the rest by MC simulation

Optimize the use of calorimeter and tracker Example: “Particle Flow” in CMS A typical high pT jet

Background subtraction and

jet clustering

Towers

Δη x Δϕ0.076 x 0.076 in barrel

Jet


Underlying Event Background

Multiple parton interaction

Large underlying event from soft scattering

Jet

Need background subtraction


Background Subtractionφ


Background Subtractionφ

1. Background energy per tower calculated in strips of η. Pedestal subtraction

Estimate background for each tower ring of constant η estimated background = <pT> + σ(pT)• Captures dN/dη of background• Misses ϕ modulation – to be improved

Background level



η

φ


η

φ

Background level



η

φ


η

φ

2. Run anti kT algorithm on background subtracted towers

Background level



η

φ


η

φ


φ

3. Exclude reconstructed jetsBackground level



η

φ


η

φ


η

φ

3. Exclude reconstructed jetsRecalculate the background energy

η

φ

Background level



η

φ


η

φ


η

φ

3. Exclude reconstructed jetsRecalculate the background energy

η

φ

4. Run anti kT algorithm on background subtracted towers to get final jets

Background level


Summary of Jet Reconstruction

Remove underlying events contribution

MC SimulationPYTHIA

Raw jet energyBackgroundsubtraction

Jet energy correction

Jet energy

correction


Flavor Creation Candidate (pp @ 7 TeV)

Reconstructed secondaryvertices from b and c quarks

Flavor Creation Candidate (pp @ 7 TeV)


-jet correlations

No -decorrelation

Increasing pT-imbalance

Less jet partners above threshold

Jets lose ~14% of their initial energy

~20% of photons lose their jet partner

PbPbPbPbPbPbPbPb

xJ=pTjet/pT

RJ = fraction of photons with jet partner >30 GeV/c

PLB 718 (2013) 773


Overlap zone is almond-shaped→ Parton energy loss is smaller along the short axis→ More high-pT tracks and jets closer to the event plane

→ Azimuthal asymmetry (v2):

→ v2 is sensitive to the path-length dependence of the energy loss

Participant plane pp

EP

Path length dependence of jet energy loss?

pT

v2

L2

L3


Jet and high pT track v2 at the LHC

• Jet and high pT track v2 : non-zero up to very high pT

• Sensitive to the path length dependence of energy loss

Jet v2

High pT track v2

PRL 109 (2012) 022301


pPb runSuccessful pPb data-taking with physics object triggers fully deployed on Sep 2012!

• The first unexpected result already came out:

Observation of long-range near-side angular correlations in proton-lead collisions at the LHC

2013 pPb run: >30/nb recorded!

• Jet quenching in pPb collisions?

• Are jets modified in pPb collisions?

• How shadowing effect and modification on the jet observables?

5x larger than pp!Elliptic flow?Color glass condensate?Modified jet structure?

Two particle correlation function

pp ridge paper: JHEP 1009 (2010) 091

PLB 718 (2013) 795


Compare photon-jet momentum balance

Xjγ = pTJet/pT

photon

in vacuum (pp collision) to the QGP (PbPb collision)

PbPb

PYTHIA

Jets lose about 15% of their initial

energy

In addition, 20% of photons lose their jet partner

(jet pT> 30 GeV/c)

PbPbPbPb

Photon-jet momentum balance

R=0.3

Xjγ

PLB 718 (2013) 773


Tracking efficiency


Jet resolution and energy scale


Subtracted background


Effects to be considered in analysis

• Impact of background subtraction• Jet response dependence on the jet event configuraiton (ex: 3-jet v.s. 2-jet

event) studied with PYTHIA & PYTHIA+HYDJET MC

• Jet flavour dependence (Quark v.s. gluon, modified jet shape and FF pattern) studied PYTHIA & PYTHIA+HYDJET MC, cross-check with PYQUEN generator

• Shape of medium response (?)• Sensitivity to tracking efficiency (and fluctuation) studied with

PYTHIA+HYDJET

• Impact of jet energy resolution• Resolution of calorimeter resolution studied PYTHIA & PYTHIA+HYDJET MC, cross-

check with PYQUEN generator

• Possible bias toward positive UE fluctuation Random cone & PYTHIA+HYDJET

• Impact to jet energy and pointing resolution PYTHIA+HYDJET

• Fake jets from UE fluctuation PYTHIA+HYDJET, data driven from dijet Δφ correlation

• Inefficiency due to downward UE fluctuation PYTHIA+HYDJET • Impact of flow and event plane dependence PYTHIA+HYDJET

• Centrality determination• Inference of the presence of a jet to centrality determination PYTHIA+HYDJET,

cross-checks with other detectors

• Detector related effects• Calorimeter noise data driven rejection studied with dijet Δφ correlation

• Fake tracks PYTHIA+HYDJET, studied with dijet Δφ correlation


Selection of b-jet Results from CMS

Identification of b-quark jets with the CMS experiment

CMS Physics Analysis Summary BTV-11-004

Inclusive b-jet production in pp collisions at √s = 7 TeV

JHEP 1204 (2012) 084, arXiv:1202.4617

Measurement of BB Angular Correlations based on Secondary

Vertex Reconstruction at √s = 7 TeV

JHEP 1103 (2011) 136, arXiv:1102.3194

Measurement of the b-jet to inclusive jet ratio in PbPb and pp collisions at √sNN = 2.76 TeV with the CMS detector

CMS Physics Analysis Summary HIN-12-003

Selection of b-jet Results from CMSb-jet results from CMS


b-jet Identification

Long lifetime of b (~1.5 ps) leads to measurable (mm or

cm) displaced secondary vertices (SV)

Subsequent charm decay may lead to a tertiary vertex

Several classes of b-jet taggers using:o Reconstructed SV’s, employing

discriminating variables such as SV mass, flight distance, etc.

o Impact parameter (IP) of tracks associated to the jet, w/o requiring a reco’d SV

o Muons in jets, exploiting the large branching ratio (20%)

b-jet identification


Bottom Production

At NLO

Excitation of sea quarks b(b) + light

dijet, w/ b(b) at beam rapidity

Gluon splitting into b and b which can

be reconstructed as a single jet

Flavor Creation (FCR) Flavor Excitation (FEX) Gluon Splitting (GSP)

LO b-b production (FCR) not dominant at the LHC

arXiv:0705.1937

pp @ 14 TeV

bottom production


Gluon Splitting Candidate (pp @ 7 TeV)gluon splitting candidate (pp @ 7 TeV)





Ncoll Number of binary scatterings

Npart Number of participating nucleons

Npart = 2 Ncoll = 1

Npart = 5 Ncoll = 6

Example:



RAA < 1 (suppression)dηdpσd

dηdpNd

N

σ=R

Tpp

TAA

coll

inelpp

AA /

/2

2 “QCD Medium”

“QCD Vacuum”

RAA > 1 (enhancement)

RAA = 1 (no medium effect) ~


inelpp

collAA σ

N=T ''NN equivalent integrated luminosity

per AA collision'‘Reduces the uncertainty from pp inclusive cross-section


‘Nuclear modification factors’

Can also be written as 1/TAA

Ncoll Averaged number of binary scattering


RHIC Lesson : Jet Quenching

Jet-quenching (Bjorken, 1982)



4< pT

trig < 6 GeV/c

pT

assoc > 2 GeV/c

Indication of jet quenching: Suppression of high p

T particles

Jet-quenching (Bjorken, 1982)


Parton energy loss models:Possible modification of jet fragmentation and broadening of the dijet ΔΦ distribution

Difficulty at RHIC:High-p

T probes are rare and direct jet reconstruction is very hard

Collisionalenergy loss

Radiativeenergy loss



Need rules to group the hadrons

A popular algorithm is anti-kT algorithm Used in ALICE, ATLAS and CMS analyses

Cacciari, Salam, Soyez, JHEP 0804 (2008) 063

Small radius parameter jet spliting

Large radius parameter

ΔR = 0.2, 0.3, 0.4, 0.5 are used in LHC analyses

Radius parameter: decide the resolution scale

Jet reconstruction


Effects to be considered in analysis

• Impact of background subtraction• Jet response dependence on the jet event configuraiton (ex: 3-jet v.s. 2-jet

event) studied with PYTHIA & PYTHIA+HYDJET MC

• Jet flavour dependence (Quark v.s. gluon, modified jet shape and FF) studied PYTHIA & PYTHIA+HYDJET MC, cross-check with PYQUEN generator

• Shape of medium response • Sensitivity to tracking efficiency (and fluctuation) studied with

PYTHIA+HYDJET

• Impact of jet energy resolution• Resolution of calorimeter resolution studied PYTHIA & PYTHIA+HYDJET MC, cross-

check with PYQUEN generator

• Possible bias toward positive UE fluctuation Random cone & PYTHIA+HYDJET

• Impact to jet energy and pointing resolution PYTHIA+HYDJET

• Fake jets from UE fluctuation PYTHIA+HYDJET, data driven from dijet Δφ correlation

• Inefficiency due to downward UE fluctuation PYTHIA+HYDJET • Impact of flow and event plane dependence PYTHIA+HYDJET

•other detectors

• Detector related effects• Calorimeter noise data driven rejection studied with dijet Δφ correlation

• Fake tracks PYTHIA+HYDJET, studied with dijet Δφ correlation

• Centrality determination• Inference of the jet to centrality determination PYTHIA+HYDJET, cross-checks with


Fraction of jets with an away side jet• Given a leading jet with pT > 150 GeV/c, >90% of them has a away side partner

• Fake away side jet rate is < 4%

Anti-kT jet R = 0.3

PLB 712 (2012) 176


Colorless Probes in Heavy Ion Collisions

ZZμμ++μμ--

μ

Wμυ

Isolated photon

γ/W/Z

QGP


Ncoll scaling confirmedin PbPb collisions at 2.76 TeV

W using single muon recoil against missing pT

Z0

arXiv:1205.6334

PRL 106 (2011) 212301CMS-PAS HIN-12-008

Isolated photon

arXiv:1205.6334PLB 715 (2012) 66

PLB 710 (2012) 256

(Non-) Suppression of Colorless Probes

R AA=σpp

inel

⟨N coll⟩

d2N AA / dpT dη

d2σ pp/ dpT dη

If PbPb = superposition of pp ...


Photon RAA

and RCP

Where Does the Energy Go?

• Suppression of high pT jets• Large dijet (photon-jet) energy (momentum) imbalance

ΔET ~ O(10) GeV,

~10% shift in <dijet pT ratio>

Where does the energy go?


arXiv:1102.1957 [nucl-ex]

0-30% Central PbPb

balanced jets unbalanced jets

Missing pT||:

Missing-pT||

Calculate projection of pT on leading jet axis and average over selected tracks with

pT > 0.5 GeV/c and |η| < 2.4

Underlying events cancels

Sum over all tracks in the event

pT

Track

pT

Track ||

ΔΦ


0-30% Central PbPb

balanced jets unbalanced jets

Missing pT||:

Integrating over the whole event final statethe dijet momentum balance is restored

excess away from leading jet

excess towards leading jet p

TTrack

pT

Track ||

ΔΦ

Balanced!!

Missing-pT||

TrackpT > 0.5 GeV/c


0-30% Central PbPb

balanced jets

Missing pT||:

The momentum difference in the dijet is

balanced by low pT particles outside the jet cone

All tracks

Tracks inthe jet coneΔR<0.8

Tracks out ofthe jet coneΔR>0.8

unbalanced jets

Inside the jet conesExcess towards leading jet

Out of the jet conesExcess towards sub-leading jet

Missing-pT||

R=0.5


Summary Need realistic MC generator (for both jet and UE)

Iterative feedback cycle is very important (like PYTHIA vs pp

data in high energy community)

CMS is willing to use and check the simulated results if you

offer a jet event generator

Need to include reconstruction effect before comparing to data

Generator level parton Energy loss + hadronization

apply jet energy smearing apply jet selection

compare the result

To Go Beyond Qualitative Observation

Generator Experiment

Used to derive correction or to compare with data

Feedback and improve the generator


Impact of Jet Energy Smearing

0-20%

Jet energy smearing

Swapped leading and subleading jet

Subleading jet is replaced by third jet

Generator level leading and subleading jets matches reco level

Dijet pT ratio Generator level jets from PYTHIAAnti-kT jet R = 0.3

Smearing function fromPLB 718 (2013) 773

yen-jie lee (cern) for the cms collaboration rencontres ion lourds/heavy ion meeting

Documents