Understanding J/ Ψ Suppression Cold Nuclear Matter (CNM) extrapolations from p(d)+A to A+A

5/25/2009 Mike Leitch 1

Understanding J/Ψ Suppression

Cold Nuclear Matter (CNM) extrapolations from p(d)+A to

A+A

Present (PPG078) CNM Constraints on A+A data

CNM effects (EKS shadowing + dissociation from fits to d+Au data, with R. Vogt calculations) give large fraction of observed Au+Au suppression, especially at mid-rapidity

more accurate d+Au constraint soon from 2008 data

d+Au

small-x(shadowing region)

PRC 77,024912(2008)

Rd

Au

5/25/2009 2Mike Leitch

& Erratum: arXiv:0903.4845Au+Aumid-rapidity

Au+Auforward-rapidity

RA

AR

AA EKS

shadowing

band

New 2008 d+Au J/Ψ data - RCP

Initial d+Au J/Ψ update from new 2008 data (~30x 2003)

• RCP pretty flat vs centrality at backward rapidity; but falls at forward rapidity (small-x)

• more soon – precision statistics requires precision systematics & careful analysis

%8860%8860

%200%200

%200

collinv

collinv

CPNN

NNR

EKS σ = 0,1,2,3,4,…15



• similar to before, use models with shadowing & absorption/breakup• but allow effective breakup cross section to vary with rapidity

• to obtain good description of data for projections to A+A

• get “breakup(y)”; compare to E866/NuSea & HERA-B• Lourenco, Vogt, Woehri - arXiv:0901.3054

• common trend, with large increasing effective breakup cross section at large positive rapidity• need additional physics in CNM model – e.g. initial-state dE/dx

with EKS shadowing

with NDSG shadowing

New CNM fits using 2008 PHENIX d+Au Rcp


Cross Check - Comparision of New Effective Breakup Cross Section fits to published 2003 d+Au RdAu Results

Fairly consistent with RdAu from old 2003 data• PRC 77,024912(2008)


Survival Probability after dividing out CNM “extrapolation”

The relation between the charged multiplicity and NPart is obtained

AuAu using PHOBOS data(Phys.Rev.C65 061901 (2002)

PbPb using NA50 data(Phys.Lett.B 530 1-4 (2002) 43-55)

Good agreement between PbPb and AuAu

Results are shown as a function of a the multiplicity of charged particles (~ energy density, assuming SPS~RHIC)


Both Pb-Pb and Au-Au seem to depart from the reference curve at NPart~200

For central collisions more important suppression in Au-Au with respect to Pb-Pb

Measured/Expected SPS results are compared with AuAu RHIC RAA results normalized to RAA(CNM)

Comparison with new RHIC results


Open Charm Nuclear Dependence from FNAL-E789/E866


Fermilab E789: D0 & B J/ψ X(charm & beauty using silicon)

Dimuon spectrometer+

16-plane, 50m pitch/8.5k strip silicon

vertex detector

upstream downstream

B J/ψ + X

D0 -> K

K+-K-+

Mass (GeV/c2)


E866/NuSea Open Charm Measurement

DumpTarget

target

dump • hadronic cocktail explains ~30% of target & <5% of dump ’s

• as expected since dump absorbs light hadrons before they can decay

• charm decays consistent between Cu target and Cu dump• use same method for Be to get nuclear dependence

beam

• data• hadrons• charm

E866/NuSea 800 GeV p+A• S. Klinksiek thesis - hep-ex_0609002• paper in preparation

2.34 m

charm ~ 3.3

hadrons

charm ~ 20

hadrons


Rapidity dependence of open charm

Open-charm p+A nuclear dependence (single- pT > 1 GeV/c) – very similar to that of J/Ψ• dominant effects are in the initial state

• e.g. shadowing, dE/dx, Cronin

• weaker open-charm suppression at y=0 attributed to lack of absorption for open charm

E866/NuSea 800 GeV p+A

PHENIX Au+Au data shows suppression at mid-rapidity about the same as seen at the SPS at lower energy• but stronger suppression at forward rapidity.• Forward/Mid RAA ratio looks flat above a centrality with Npart = 100

Several scenarios may contribute:• Cold nuclear matter (CNM) effects

• important, need better constraint

• Sequential suppression• QGP screening only of C & ’- removing their feed-down contribution to J/ at both SPS & RHIC

• Regeneration models• give enhancement that compensates for screening

PHENIX A+A Data and Features

Centrality (Npart)


Looking for the cold nuclear matter baseline for J/ψ production at RHIC

Tony FrawleyFlorida State University

ECT, TrentoMay 26, 2009

May 26, 2009 Tony Frawley, FSU 16

Many thanks for contributions/help from:

Ramona Vogt

Mike LeitchAlex Linden LevyJamie NagleDarren McGlinchey


The quarkonium plan

Species Purpose

p+p Quarkonium production mechanismsBaseline cross sections for heavy ions

d+Au Cold nuclear matter effectsBaseline CNM R

AA for heavy ions

Cu+Cu Hot nuclear matter effects near TC

Au+Au Hot nuclear matter effects well above TC


The PHENIX Detector

J/ψ→e+e-

-0.35 < y < 0.35J/ψ→μ+μ-

1.2 < |y| < 2.2


Brief review of the relevant J/ψ data

Au+Au RAA

Run 4 Au+Au + Run 5 p+p

Cu+Cu RAA

Run 5 Cu+Cu + Run 5 p+p

d+Au RCP

Run 8 d+Au

To come: Run 8 RdAu

with Run 6 pp reference


Reference data – Run 5 p+p

PHENIX, PRL98, 2002 (2007)This is the reference data set for all nuclear modification factors shown here.


Cu+Cu and Au+Au RAA

Phys. Rev. Lett. 101, 122301 (2008)The Npart dependence of

Au+Au and Cu+Cu is consistent.

Note the smaller systematic uncertainties for the Cu+Cu data. This is primarily due to smaller uncertainties on Ncoll from the Glauber calculation.

Thus the Cu+Cu data will be much better for studying the onset of hot nuclear matter effects.


Au+Au RAA

PHENIX – reference here

The stronger Au+Au suppression at forward/backward rapidity has generated considerable interest.

But what is the expected suppression due to cold nuclear matter effects?


d+Au RCP

The first results for d+Au from Run 8, shown at QM09.

Four centrality bins to make three R

CP points:


What can we learn from the existing charmonium RAA

data?

The heavy ion charmonium data alone have not taught us as much as we would like, because of serious uncertainties caused by:

1) Poorly known initial state effects at RHIC: Break up cross section for collisions with nucleons. Shadowing. Other effects? Initial state energy loss?

2) Poorly known open charm production cross sections.

Thus the trade-off between coalescence and destruction is difficult to illuminate experimentally.

To try to make inroads on 1), we start from the most recent d+Au data set: – Run 8 d+Au

First we briefly review previous attempts to use Run 3 d+Au data for this.


Estimating the CNM RAA

from Run 3 d+Au data - 1

This has been done before in three ways:

1) PHENIX (Phys. Rev. C 77, 024912 (2008) and erratum arXiv:0903.4845) fitted Run 3 R

dAu using a

single σbreakup

at all rapidities + EKS98/nDSg

shadowing calculations by Ramona Vogt.

The CNM RAA was estimated using calculations of R

AA for Cu+Cu and Au+Au by Ramona, using the

fitted σbreakup

+ EKS98/nDSg.

However (Phys. Rev. Lett. 101, 122301 (2008)PPG071), when a single σ

breakup is used at all

rapidities the ratio of the predicted y=0 to |y|=1.7 CNM RAA values is just a prediction of the shadowing model. Therefore this is not a good way to use the R

dAu data to test if the increased suppression at

|y|=1.7 is due to CNM effects.R

AA

Au+Aumid-rapidity

Au+Auforward-rapidity

RA

A

EKS shadowi

ngband




2) Raphael Granier de Cassagnac (J. Phys. G34, S955 (2007)) used direct folding of the R

dAu data, with some assumptions, to predict

the CNM RAA

for Au+Au. This works only for

Au+Au, since RdAu

is used directly.

This approach produces completely independent CNM R

AA values at y=0 and |y|

=1.7 – which is very good. But because of the low statistical precision of the Run 3 d+Au data, the results are inconclusive.

This approach cannot be used with d+Au RCP

data, nor can it be used to estimate a CNM R

AA baseline for Cu+Cu.




Phys. Rev. Lett. 101, 122301 (2008)

3) The PHENIX RdAu

data were fitted

separately at y=0 and |y|=1.7 with σbreakup

+ EKS98/nDSg calculations by Ramona. The CNM R

AA was predicted for Cu+Cu

and Au+Au independently at y=0 and |y|=1.7 using calculations by Ramona.

While this makes the ratio of the estimated CNM R

AA at y=0 and |y|=1.7 sensitive to

the RdAu

data, it still assumes the forward

and backward rapidity data have the same σ

breakup. We will see this is not justified.

NOTE: The uncertainty bands here are underestimated due to the fitting error that was corrected for 1). Not fixed here yet!


Fitting the Run 8 d+Au RCP

We want to parameterize the d+Au RCP

data so that we can predict the heavy ion

RAA

that would result from p+A physics only.

Fit RCP

vs centrality independently at each rapidity using calculations of RdAu

vs

impact parameter by Ramona Vogt that include: σ

breakup for collisions of (forming) J/ψ with nucleons (0-15 mb, 1 mb steps).

A shadowing model – EKS98 and nDSg are used here.

Convert RdAu

vs impact parameter to RdAu

vs centrality using PHENIX Glauber

impact parameter distribution for each dAu centrality bin.

Fit procedure: Fit R

CP vs centrality using only uncertainties that are uncorrelated in rapidity.

Vary RCP

by +/- 1σ in uncertainties that are correlated in rapidity, and refit. Vary R

CP by +/- 1σ in uncertainties that are global with rapidity and refit.

Uncertainties are shown respectively as bars, boxes, and a global number.


Fits to d+Au RCP

– example for EKS98

Integrated for each muon arm


σbreakup

vs y from d+Au RCP

fits with EKS98 and nDSg


Comparison with lower energy data – EKS98 fits

Lourenco, Vogt and Woehri (JHEP 02 (2009) 014) published the effective breakup cross section vs y from fits to E866 and HERA-B data.

Our results from 200 GeV are shown here compared with their results for the EKS98 case.

For y > 1.2 the 200 GeV data follow the trend observed at lower energy remarkably closely!


Comparison with lower energy data – nDSG fits

Note that the effective breakup cross section is significantly lower for y < 1.2.

But for y > 1.2 there is little difference from the EKS case.


Sanity check! Comparision of new effective breakup cross section fits from R

CP to published 2003 d+Au RdAu

results

From the talk by Mike Leitch.

Fairly consistent with centrality integrated RdAu from old 2003 data• PRC 77,024912(2008)


Parameterize d+Au RCP

at |y|= 0, 1.7 – EKS98

Integrated for each muon arm


Effective σbreakup

used in Glauber calculations |y| = 0, 1.7


Cold Nuclear Matter RAA

for heavy ions

Having “calibrated” the Vogt calculations at each rapidity, we estimate the CNM R

AA using the results from the dAu R

CP fits.

To do this, we use a Glauber calculation for Au+Au that reproduces well the average Npart and Ncoll values for the centrality bins used by PHENIX.

In the Glauber calculation:Each nuclear collision is placed in a centrality bin according to Npart. For each nucleon-nucleon collision: Determine impact parameter b1 of nucleon 1 in its target nucleus. Determine impact parameter b2 of nucleon 2 in its target nucleus. Add to the accumulated RAA: RdAu(b1,y=0) * RdAu(b2,y=0) Add to the accumulated RAA: RdAu(b1,y=-1.75) * RdAu(b2,y=1.75)

After processing all events, print out at y=0 and y=1.7 for centrality bin j: Nevts[j], Σ(RAA[j])/Nevts[j], Σ(Ncoll[j])/Nevts[j], Σ(Npart[j])/Nevts[j]


Estimation of uncertainties

The CNM RAA

is calculated for the central fitted σbreakup

and for +/- 1σ in the

type A (uncorrelated in rapidity) uncertainty and for +/- 1σ in the type B (correlated in rapidity) uncertainty.

The type A uncertainty is shown as a vertical bar, and the type B as a box.


Heavy ion CNM baseline RAA

– EKS98 parameterization


Heavy ion CNM baseline RAA

– nDSg parameterization


Calculating the heavy ion RAA

“survival probability”

Now we can calculate the ratio RAA

/RAA

(CNM) from the measured RAA

and

the estimated RAA

(CNM) shown on the previous slides.

In the following plots the uncertainty in RAA

/RAA

(CNM) due to the

uncorrelated (mostly statistical) uncertainty in the measured RAA

is shown

as a bar, the correlated uncertainty in the measured RAA

us shown as a

narrow box, and the uncertainty due to the estimated CNM RAA

is shown

as a wider box.


Heavy ion “survival probability” at y=0 (EKS example)


Heavy ion “survival probability” at |y| = 1.7 (EKS example)


Heavy ion “survival probability” - EKS98 parameterization


Heavy ion “survival probability” - nDSg parameterization


Comments

We are just starting to try to understand the d+Au data and their implications for heavy ions. Keep several things in mind about what was done here:

We assume that we can isolate hot nuclear matter effects by calculating RAA

due to

a (Glauber guided) superposition of d+Au collisions. Perhaps not!

The role of Glauber uncertainties (mostly Ncoll) needs to be understood in detail. The systematic uncertainties should be considered tentative until then.

We believe that fitting the RdAu

data, rather than RCP

, will provide greater precision

when estimating the CNM baseline RAA

for Au+Au and Cu+Cu.

Although the parameterization at each rapidity is precise, it would be more satisfying if the model worked well over the full d+Au rapidity range!

We need to repeat this exercise for the Cu+Cu data (easy enough to do, did not have time yet).


Summary

The PHENIX d+Au data at 200 GeV seem to follow the trend observed at lower energy of a rapid rise in the effective σ

breakup at forward rapidity.

The effective σbreakup

appears to be roughly constant below y ~ 1.25 at 200 GeV.

The RAA

(CNM) estimated from the fits to the RdAu

data show significantly stronger

suppression at |y|=1.7 than at y=0.

The measured suppression beyond the estimated RAA

(CNM) values, presumably

due to hot nuclear matter effects, seems to be very similar at y=0 and |y|=1.7 at about 50%.


What next

Calculate RAA

(CNM) for Cu+Cu.

Investigate the transverse momentum dependence.

Understand the role of Ncoll uncertainties better.

Do it all again with RdAu

instead of RCP

.


Backup


EPS08 parameterization


Existing RHIC Data - Au+Au


Existing RHIC Data - Cu+Cu


Reference data: Run 6 p+p will be used for RdAu

The best p+p data set that has been analyzed so far is from Run 6.

Understanding J/ Ψ Suppression Cold Nuclear Matter (CNM) extrapolations from p(d)+A to A+A

Documents

Transcript of Understanding J/ Ψ Suppression Cold Nuclear Matter (CNM) extrapolations from p(d)+A to A+A