Charged-particle dN/d from PHOBOS

A. H. Wuosmaa

(Argonne National Laboratory)

for the

PHOBOS Collaboration

Quark Matter 2001

The PHOBOS CollaborationARGONNE NATIONAL LABORATORY

Birger Back, Nigel George, Alan Wuosmaa

BROOKHAVEN NATIONAL LABORATORY

Mark Baker, Donald Barton, Alan Carroll, Stephen Gushue, George Heintzelman, Robert Pak, Louis Remsberg, Peter Steinberg, Andrei Sukhanov

INSTITUTE OF NUCLEAR PHYSICS, KRAKOW

Andrzej Budzanowski, Roman Holynski,, Jerzy Michalowski, Andrzej Olszewski, Pawel Sawicki , Marek Stodulski, Adam Trzupek, Barbara Wosiek, Krzysztof Wozniak

MASSACHUSETTS INSTITUTE OF TECHNOLOGY

Wit Busza* , Patrick Decowski, Kristjan Gulbrandsen, Conor Henderson, Jay Kane , Judith Katzy, Piotr Kulinich, Johannes Muelmenstaedt, Heinz Pernegger, Corey Reed, Christof Roland, Gunther Roland, Leslie Rosenberg, Pradeep Sarin, Stephen Steadman, George Stephans, Gerrit van

Nieuwenhuizen, Carla Vale, Robin Verdier, Bernard Wadsworth, Bolek Wyslouch

NATIONAL CENTRAL UNIVERSITY, TAIWAN

Willis Lin, JawLuen Tang

UNIVERSITY OF ROCHESTER

Erik Johnson, Josh Hamblen, Nazim Khan, Steven Manly, Inkyu Park, Wojtek Skulski, R. Teng, Frank Wolfs

UNIVERSITY OF ILLINOIS AT CHICAGO

Russell Betts, Clive Halliwell, David Hofman, Burt Holzman, Wojtek Kucewicz, Don McLeod, Rachid Nouicer, Michael Reuter

UNIVERSITY OF MARYLAND

Richard Bindel, Edmundo Garcia-Solis, Alice Mignerey

* spokesperson

Why study dNch/d?• dNch/d is sensitive to all aspects of

charged-particle production in heavy-ion collisions:– Interplay between hard and soft processes– Effects of shadowing, jet quenching– Reaction dynamics, re-scattering– Full distribution reflects a time integral of

particle production throughout the collision and total entropy production

• Lots of existing data for pp, pA, AB,AA– How do RHIC data fit into this picture?

Charged-particle Multiplicity in Au-Au at RHIC Energies

Influence of various physical effects on dNch/din very central collisions

(Wang & Gyulassy,Private communication)

130GeV)NNs(

Octagon Multiplicity Detector

Ring Multiplicity Detectors

The PHOBOS Detector

All multiplicity detectorsare silicon pad sensors

Coverage of multiplicity detectors in

0 1 2 3 4 512345

1m2m

5m

Schematic Multiplicity Detector

coverage for vertex at Z=0.

PHOBOS Multiplicity detectors

Octagon 1 of 6 Rings

What I will not discuss:

• Event Selection

• Vertex Determination

• Centrality DeterminationSee Talks by:

R. Pak, Tuesday 3:20 J. Katzy, Wednesday 4:40

And see Poster by: P. Decowski

1 Count hits binned in , centrality (b)

2 Calculate acceptance A(ZVTX) for that event

3 Find the occupancy per hit pad O(,b)

4 Fold in a background correction factor fB(,b)

dNch

d =hits

O(,b) ×fB(,b)

A(ZVTX)

E depositionin multiplicitydetectors for 1 event.

“Measuring” the occupancy

!)(

N

eNP

N

N=number of tracks/pad=mean number of tracks/pad

The numbers of empty, and occupied, padsdetermine the occupancy as a function of ,b

Method: Assume Poisson statistics

Ntr

acks

/hit

pad

0-3%

50-55%

Octagon

Rings(central)

(peripheral)

Discriminating background with E E

(“M

IP”)

20 64-2-6 -4

04

81

2

20 64-2-6 -4

E (

“MIP

”)0

48

12

Data Monte Carlo

Si

E vs. in the OctagonFrom vertex

Not from vertex

MC, Occupancy corrected

MC “truth”

Compare PHOBOS Monte Carlo “data” analyzed usingoccupancy corrections to “truth” - the difference gives corrections for remaining background.

f B(

,b)

fB=MCTruth/MCOcc

dNch

/d

Estimating remaining

backgrounds

-6 -4 -2 0 2 4 6

-6 -4 -2 0 2 4 6

0.2

0.4

0.6

0.8

1.0

200

400

600

45-55% 35-45% 25-35%

15-25% 6-15% 0-6%

dNch

/d

dNch

/d

dNch/d for different centrality bins

Octagon RingsStatistical Unc. onlyStatistical Unc. only PHOBOS Prelim.

Centrality Dependence of Nch(||<5.4)

Npart

PHOBOS Prelim.

±10% Systematic Uncertainty±10% Systematic Uncertainty

HIJING

Nch

(||

<5.

4)

Shapes of dNch/d for different Npart

dNch

/d

(dN

ch/d

)/(

½N

pa

rt)

dNch

/d

Data

HIJING

HIJING

(dN

ch/d

)/(

½N

pa

rt)

Systematic error ±(10%-20%)Systematic error ±(10%-20%)PHOBOS Prelim.

354

216

102

Mean Npart% 0-3

15-20

35-40

Data

Centrality dependence of dNch/d|

Npart

(dN

ch/d

)/(

½N

pa

rt)

||

<1

5-5.4

4-4.4

3-3.4

2-2.4

PHOBOS Prelim.Symbols: Solid lines: HIJING

Errors are systematic

Summary• First multiplicity distributions over 4 now

available at for wide range of impact parameters

• Nch(||<5.4)=4100±410 for the 3% most central collisions

• Distributions are somewhat wider than predicted by some models

• (dNch/d/(½Npart) in fragmentation region drops by ~½ from Npart=100 to 350

• Outlook: coming analyses: EbyE, d2N/dd(I. Park, Wed. 5:55)

130GeVNNs

“The appetizers were truly delightful, and Iam anxiously looking forward to the main course” -N. Khrushchev

Or -

We can’t wait for 200 GeV! NNs

Backup slides to follow:

Data and Monte Carlo

DataMonte Carlo

Comparison of DE measuredin the Octagon at ||~3.0, with spectrum predicted by Monte-Carlo:Conclusion: MC is a good predictor of background.

(dN

ch/d

)/(

½N

pa

rt)

Backup Method: Study E spectra

Ntr

acks

/hit

pad

E (MIPs)

Cou

nts

1 32

1hit

2hits3hits

oct rings

Occupancy reality check

Can determine from DE relative yields of events with1,2,3… hits/pad: gives independent measurement of Occupancy

Good: fewer potentially unjustified assumptionsBad: Fitting procedure somewhat unstable, occasional large errors

Very Good: We get the same answer in the end!!

Comparison of all methods at mid rapidity

MULT ARRAY

d

d n

n

1

)GeV(s24 31 5345 63

2-4

5-9

10-1

415

-19

20-2

4T

otal

obs

erve

d m

ultip

licity

pp data from ISR

W. Thome et al.,Nucl. Phys. B129(1977) 365.