open charm production at

Open Charm ProductionOpen Charm Productionatat . .

Andrew Glenn

University of Tennessee

July 7, 2004

Andrew Glenn2

OutlineOutline

• Motivation (Experimental and Theoretical)

• Single muon Au+Au analysis• Other charm data at PHENIX and STAR

• Summary and Outlook

Andrew Glenn3

Why is Charm Interesting (in A+A)?Why is Charm Interesting (in A+A)?Hadronic

Vs.

QGPCan charm help show the difference?

Pre-thermal

Thermal

Initial production

Andrew Glenn4

d[A+BX] =

ij f

i/Af

j/B d [ijcc+X] D

cH + ...

J.C.Collins,D.E.Soper and G.Sterman, Nucl. Phys. B263, 37(1986)

Factorization theorem for charmed hadron production

fi/A,

fj/B

: distribution fuction for parton i,jD

cH : fragmentation function for c

d [ijcc+X] : parton cross section

+ ... : higher twist (power suppressed by

QCD/m

c, or

QCD/p

t if p

t ≫m

c ) :

e.g. "recombination" PRL, 89 122002 (2002)

Charm production at RHIC is dominated by gluon fusion

Charm ProductionCharm Production

B. L. Combridge et al. Nuc. Phys., B151:429–456, 1979

Andrew Glenn5

Measure Thermalization Time?Measure Thermalization Time?

Initial estimates showed pre-equilibrium charm production to be similar to the initial productionin a QGP at RHIC energy.

Berndt Muller and Xin-Nian Wang. Phys. Rev. Lett., 68:2437–2439, 1992.

Andrew Glenn6

Probably not.Probably not.Refined estimates showed thepost initial production to be small.

Most charm is made via gluon fusion in the initial stage.

Peter L´evai, Berndt Muller, and Xin-Nian Wang. Phys. Rev., C51:3326–3335, 1995.

Zi-wei Lin and M. Gyulassy. Nucl. Phys., A590:495c–498c, 1995.

Andrew Glenn7

Charm Enhancement?Charm Enhancement?

One early interpretation of the NA45 electron pair enhancement was a large charm enhancement

P. Braun-Munzinger, D. Miskowiec, A. Drees, and C. Lourenco. Eur. Phys. J., C1:123–130, 1998.

Andrew Glenn8

Quite Possibly (but less)Quite Possibly (but less)

NA50 Muon pair data showed that charm was not so largely enhanced. BUT there was still a significant excess which may be due to charm enhacement

Open CharmEnhancement ?

M. C. Abreu et al. Eur. Phys. J., C14:443–455, 2000.

Andrew Glenn9

Centrality DependenceCentrality Dependence

Enhancement up to 3.5 in central Pb+Pb

NA50

M. C. Abreu et al. Eur. Phys. J., C14:443–455, 2000.

Andrew Glenn10

HadronizationHadronization

fragmenting parton:ph = z p, z<1

recombining partons:p1+p2=ph

Fragmentation Vs. Coalescence

Quarks can combine with unrelated

Quarks can combine with unrelated quarks. Observed since ’77 as ‘leading particle effect’ (recombination). How much larger of a role might this play at RHIC; especially in a ‘quark soup’ (coalescence).

Andrew Glenn11

Enhancement at RHIC Vs. SPSEnhancement at RHIC Vs. SPS

Estimates show that RHIC will has a lower fraction of c quark pairs below 2mD threshold to contribute to enhancement via coalescence.

This region can’t contributeto open charm throughfragmentation.

Ma

xim

um

En

ha

nc

em

en

t

A. P. Kostyuk, M. I. Gorenstein, and W. Greiner. Phys. Lett., B519:207–211, 2001.

Andrew Glenn12

D RatiosD Ratios

• The ratios of different D mesons will differ in hadron vs. QGP scenario.

• Free relativistic fermion quark gas at T=200 MeV and a baryon chemical potential, µ = 0: D−

s / D− = 0.94

• A equilibrated hadronic bose gas at T=180 MeV: D−

s / D− = 0.610

• Final state interactions such as D± + K± → D±s +

π± could modify the ratios and need to be corrected for.

Cheuk-YinWong. Nucl. Phys., A630:487–498, 1998. and private communication

Andrew Glenn13

Other Information from CharmOther Information from CharmElectrons from charm

Transverse momentum spectra and flow parameters are sensitive to QGP.

V. Greco, C. M. Ko, and R. Rapp. Quark coalescence for charmed mesons in ultrarelativistic heavy-ion collisions. 2003.

Andrew Glenn14

Cronin

Nuclear effects such as shadowing, energy loss, kT broadening, cronin enhancement, and gluon saturation (Color Glass) need to be considered

Nuclear EffectsNuclear Effects

CGC

Shadowing

Dmitri Kharzeev and Kirill Tuchin. hep-ph/0310358, 2003.B. Z. Kopeliovich, J. Nemchik, A. Schafer, and A. V. Tarasov. Phys. Rev. Lett., 88:232303, 2002.R. Vogt. Int. J. Mod. Phys., E12:211–270, 2003.Jens Ole Schmitt et al. Phys.Lett., B498:163–168, 2001.

Andrew Glenn15

Charmonium ReferenceCharmonium ReferenceIn the past, J/ψ production (supression) has been measured relative to Drell-Yan. In order to minimize nuclear effects, such as gluon shadowing, open charm will be a better reference.

SPS/AGSRHIC

J/Psi to DY

Charmonium has a much different production mechanism than Drell-Yan.

J/Psi to Open Charm

H. Satz and K. Sridhar. Phys. Rev., D50:3557–3559, 1994.

Andrew Glenn16

The ApparatusThe ApparatusDetector (Iarocci Tubes)

Steel Absorber

Andrew Glenn17

Run II Au+Au Data SampleRun II Au+Au Data Sample• RHIC Run II Au+Au

(The first muon capable run)• Start With 7.7M Minimum-Bias Events

(Cleanest section of running period)

• Cuts on:• Vertex (-20 to 38cm)

• Track Quality (2/DOF < 7, NTrHits >=12) (-1.51 to -1.79 or = 155 -161o)

Uniform acceptance in event vertex.• Azimuthal Cut

Andrew Glenn18

Run IV Au+AuRun IV Au+Au

• Better Shielding (Tunnel and MuID Hole)

• Better Statistics (~1.5 Billion Min-bias!)

• Better Hardware Acceptance (stable HV..)

Run II

Run III

Andrew Glenn19

Sources of Muon CandidatesSources of Muon Candidates

• D (B) meson semi-leptonic decays(D K …) Prompt Signal

• Decay muons from hadron decays (±± , K±± )

• ‘Hadron’ Punch Through(±, K±, p)

• Decays from J/, , Drell-Yan…(Small Contribution)

PH

MuID

Bac

kgro

un

d

Gap 0 1 2 3 4

Steel absorber

Andrew Glenn20

Muon Identification Muon Identification

Main muon identification is from momentum/depth matchingPHMuID

ste

el

Last Gap = 2Centrality > 20%

Long tail from interacting hadrons(more evidence to come)

Sharp Muon Peak

Data –RedSimulation –Black

Gap (Plain) 0 1 2 3 4

3 GeV

1 GeV

3 GeV

Andrew Glenn21

Run IV Stopping PeaksRun IV Stopping Peaks

Muon Peaks

NorthSouth

Andrew Glenn22

We want to separate the contribution from prompt muon production and /K decays

Collision Point

nosecone

magnet

Muon trackerMuon ID

40 cm

• D c = 0.03 cm Decays before absorber• c = 780 cm Most are absorbed, but some decay first • K c = 371 cm Most are absorbed, but some decay firstγcτ >> 80cm → Decay Probability nearly constant between nosecones

Collisions occurring closer to the absorber will have fewer decay contributions. Should see a linear increase in decay background with increasing vertex.

Decay ContributionDecay Contribution

X

Z

Andrew Glenn23

Vertex DependenceVertex Dependence

=Centrality > 20%

Very Linear Shape Due to Decays

Not due to vertex detector resolution Centrality > 20%

Centrality > 20%

1 < pT > 3 GeV/c

Andrew Glenn24

Interacting Hadron Vertex Interacting Hadron Vertex

Interacting Hadrons: Last Gaps 2 and 3, Pz St3 tail. Flat shape indicates little to no decay (hence muon) component

Last Gap = 2

Centrality > 20%

Centrality > 20%

Andrew Glenn25

From simple (event vertex corrected) subtraction of near muons from far muons

(Hence NOT NORMALIZED, also NOT ACCEPTANCE/EFFICIENCY CORRECTED)

Free Decay pFree Decay pTT Distribution Distribution

Centrality > 20%

Region II – Region I

Z vertex

Muon Vertex

Region I Region II

Decays

prompt decay + punchthrough +??

Z absorber

Scaled

Andrew Glenn26

SimulationsSimulations

• With the availability of BRAHMS and K data at Muon Arm rapidities, we can examine a data driven simulation approach.

• An accurate modeling of K and (and p) production for simulation input is required for punchthrough estimations.

Decays Z dependence, simulation

Punch thoughPz Tails (extrapolation to Gap4), simulations

Promt Signal Simulations for acc*eff

Andrew Glenn27

Data Driven Particle GeneratorData Driven Particle Generator

Use scaled PHENIX central arm data to for basis of event generator. BRAHMS preliminary data helps with scaling and justification ( P(y,pT) ≈ P(y)P(pT) ). Only measurements for 5% most central events are available from BRAHMS.

5% Most Central Events

BRAHMS data extractedfrom Djamel Ouerdane’s thesis

BRAHMS5% most central

Provides scaling

Andrew Glenn28

Early Simulation ResultsEarly Simulation Results• 900K Single K’s and ’s thrown with BRAHMS (preliminary, 5% most

central) pT and dN/dy shapes. pT > 1 GeV required. Passed through full simulation/reconstruction.

Last Gap Reached Reconstructed Decay Muons Reconstructed non-muons

2 161 250

3 297 169

4 712 74

Last Gap = 4Same cut as data

Decays Punchthrough (non-muons)

Projected decay contribution ends at ~-73cm

Last Gap = 4Same cut as data

Andrew Glenn29

Rough Component BreakdownRough Component Breakdown

Free Decays

Other Decays

Prompt muons + unresolved background (combinatoric …)

Punch Through

z=-40cm nose cone

Z=-73cm decay sim.

Andrew Glenn30

Completing the AnalysisCompleting the Analysis

• Complete large statistics simulations to more accurately estimate punch though and background.

• Complete a more accurate estimate of combinatory background.

• Bin analysis in pT to determine prompt muon spectra.

• Finish acceptance*efficiency corrections.

Andrew Glenn31

Green band = systematic uncertainty.Blue dotted line = expectation.

Blue line = statistical uncertainty.

μ -

No absolute value yet; final efforts in progress

Measured decay muons with the expectation

Prompt muons

p+p Muon Open Charmp+p Muon Open Charm

Y. Kwon et. al

Andrew Glenn32

p+p Muon Open Charm IIp+p Muon Open Charm II

Work in progressestimates for backgroundsas a function of pT

Max BG

Min BG

Y. Kwon et. al

Andrew Glenn33

Has J/Has J/ψψ Muon Data Muon Data

Andrew Glenn34

J/J/ψψ Visible in Run IV Au+Au Data Visible in Run IV Au+Au Data

South ArmCentrality >40

Andrew Glenn35

PHENIX PRELIMINARY

Electron MeasurementsElectron Measurements

p+pAu+Au

consistant with binary scaling

Andrew Glenn36

.. Electron Measurements II Electron Measurements II

(e++e–)/2sys. error

Min. bias at Au+Au sNN=200GeV

PHENIX

• Fully corrected spectrum– Acceptance & eID efficiencies– 2.5M and 2.2M events

analyzed with and without the converter

• Backgrounds from KeX (less than 5%) and , , ee (1%) decays subtracted

• Systematic error is 13% at high pT

eID, acc, etc 11.8%K- >eX, VM- >ee 2.0%Signal extraction 5.0%

Andrew Glenn37

Mass plots from dAu data using event-mixing technique

D±

7.4<pt<9.3 GeV/c

D0+D0

0 < pT < 3 GeV/c, |y| < 1.0

d+Au minbias

8.5< pt<11.0 GeV/c

00 DD

STAR has measured charm in d+AuSTAR has measured charm in d+AuQM 2004

STAR also has a single electron charm measurement.

Andrew Glenn38

Summary and OutlookSummary and Outlook• Open (and hidden) charm measurements are

very important to RHIC HI/QGP physics.• PHENIX is capable of making charm

measurements via semileptonic decays (at forward and central rapidity).

• Electron measurments exhist, and muon measurements are close.

• Quicker simulations (using staged cloning) are being tested to aid in estimating punch through.

• Future upgrades may enhance our ability to do these measurements. (displaced vertex detection, pad chambers …)

Andrew Glenn39

Additional Slides

Andrew Glenn40

Peripheral Data Vs SimulationPeripheral Data Vs Simulation

Simulation: Muons From Central Hijing Data: Centrality > 60 (For cleaner events)

Falsely extended tracksLast Gap = 2

Last Gap = 3

Last Gap = 4

No clear peak or tailsince last gap.

Andrew Glenn41

Hadronization AnimationsHadronization Animations

C

U

UCUC