Russell Betts UIC

1
Multiplicity Measurements withThe PHOBOS Detector
Russell Betts (UIC) for the PHOBOS Collaboration
18th Winter Workshop on Nuclear Dynamics Nassau,
Jan 20th-27th,2002
2
The PHOBOS Collaboration
Birger Back, Nigel George, Alan Wuosmaa
Mark Baker, Donald Barton, Alan Carroll,
Joel Corbo, Stephen Gushue, George Heintzelman,
Dale Hicks, Burt Holzman,Robert Pak, Marc
Rafelski, Louis Remsberg, Peter Steinberg, Andrei
Sukhanov Andrzej Budzanowski, Roman
Holynski, Jerzy Michalowski, Andrzej Olszewski,
Pawel Sawicki , Marek Stodulski, Adam Trzupek,
Barbara Wosiek, Krzysztof Wozniak Wit
Busza (Spokesperson), Patrick Decowski, Kristjan
Gulbrandsen, Conor Henderson, Jay Kane , Judith
Katzy, Piotr Kulinich, Johannes Muelmenstaedt,
Heinz Pernegger, Michel Rbeiz, 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 Chia Ming Kuo, Willis Lin, Jaw-Luen
Tang Joshua Hamblen , Erik Johnson, Nazim
Khan, Steven Manly,Inkyu Park, Wojtek Skulski,
Ray Teng, Frank Wolfs Russell Betts,
Edmundo Garcia, Clive Halliwell, David Hofman,
Richard Hollis, Aneta Iordanova, Wojtek Kucewicz,
Don McLeod, Rachid Nouicer, Michael Reuter, Joe
Sagerer Richard Bindel, Alice Mignerey
ARGONNE NATIONAL LABORATORY BROOKHAVEN NATIONAL
LABORATORY INSTITUTE OF NUCLEAR PHYSICS,
KRAKOW MASSACHUSETTS INSTITUTE OF
TECHNOLOGY NATIONAL CENTRAL UNIVERSITY,
TAIWAN UNIVERSITY OF ROCHESTER UNIVERSITY OF
ILLINOIS AT CHICAGO UNIVERSITY OF MARYLAND
3

• Completed Spring 2001
• 4p Multiplicity Array
• - Octagon, Vertex Ring Counters
• Two Mid-rapidity Spectrometers
• TOF wall for High-Momentum PID
• Triggering
• Scintillator Paddles
• Zero Degree Calorimeter

137000 Silicon Pad channels
4

• Outline of Talk
• Centrality Determination Nparticipant and
  Ncollision
• Techniques for Multiplicity Measurements
• Tracklets
• Hit Counting
• Energy Deposition
• Results
• Energy Dependence for ???? 1
• Centrality Dependence
• dN/d? Shapes
• Summary and Taster of Future Delights

5
Triggering on Collisions
Positive Paddles
Negative Paddles
ZDC N
ZDC P
Au
Au
PN
PP
Paddle Counter
ZDC Counter
Valid Collision

• Coincidence between Paddle counters at Dt 0
  defines a valid collision.
• Paddle ZDC timing reject background.
• Sensitive to 973 of inelastic cross section
  for AuAu.

6
Trigger Selection - ZDC vs Paddles
Peripheral
7
Determining Centrality
Counts

• HIJING GEANT
• Glauber Calculation
• Model of Paddle Response

Paddle signal (a.u.)
Counts
Npart
8
Uncertainty on Npart

• Measurement sensitive to trigger bias
• Minimum-bias still has bias
• Affects most peripheral events

Counts

• Estimating 97 when really 94 overestimates Npart

Paddle signal (a.u.)
9
Multiplicity Distributions
Hits in One Layer of Silicon
Rings
Vertex
Octagon
10
AuAu Collision Event Display
11
Event Vertex Finding
z

• Vertex Resolution
• sx 450 mm
• sy sz 200 mm

12
Vertex Tracklet Reconstruction
Tracklets are two point tracks that are
constrained by the event vertex.
dh h1 h2 df f1 f2
dh lt 0.04 df lt 0.3
13
Combinatorial Background
All Pairs of Hits
Background Flip
Outer Hit Bin 10 (Data)
14
Backgrounds
Weak Decays
d Electrons
15
Vertex Tracklet Systematic Error

• Reconstruction Vertex selection, Tracklet
  algorithm etc.
• 1.8

• Weak Decays Mostly Ks and L - 2

• Background Combinatorial, d-electrons - 1.5

• MC Generators Different particle production,
  background etc.
• - 5

• Total 7.5

16
Analog and Digital Hit-Counting
f
h
Octagon, Ring and Vertex Detectors
(unrolled) Count Hits or Deposited Energy
17
Discriminating Background with dE
Monte Carlo
Data
DE (MIP)
h
h
Not from vertex
Si
DE vs. h in the Octagon
From vertex
18
DE deposition in multiplicity detectors for 1
event.
f
h

• Count hits binned in h, centrality (b)
• Calculate acceptance A(ZVTX) for that event
• Find the occupancy per hit pad O(h,b)
• Fold in a background correction factor fB(h,b)

19
Measuring the Occupancy
Method Assume Poisson statistics
0-3
Octagon
(central)
Rings
Ntracks/hit pad
Nnumber of tracks/pad m mean number of
tracks/pad
50-55
(peripheral)
h
The numbers of empty, and occupied,
pads determine the occupancy as a function of h,b
20
Estimating remaining backgrounds
MC, Occupancy Corrected
600
dNch/dh
400
MC truth
fBMCTruth/MCOcc
200
h
Compare PHOBOS Monte Carlo data analyzed
using occupancy corrections to truth - the
difference gives corrections for remaining
background.
fB(h,b)
h
21
Energy Loss ? Multiplicity
300 mm Si
Energy deposited in ith pad (truncated) corrected
for angle of incidence
Mean energy loss for one particle traversing pad
RATIO OF TOTAL TRACKS TO PRIMARY TRACKS
0.30 - 0.40
Measured S/N 10 - 20 ltlt Landau Width Use
Non-Hit pads - for Common-Mode Noise
Suppression M 240 15 5 CMN for one sensor
(120 channels) at h 0
22
Uncertainty in Theoretical Predictions
23
Constraining the Models
24
Ratio 200/130 GeV
Ratio 200/130 averaged for four PHOBOS methods
Phobos Measurement
R200/130 1.14 /- 0.05 Moderate Increase in
Energy Density?
Systematic Uncertainty
25
Hard and Soft Processes

• Soft processes (pT lt 1 GeV)
• Color exchange excites baryons
• Baryons decay to soft particles
• Varies with number of struck nucleons
• Wounded Nucleon Model
• Hard processes (pT gt 1 GeV)
• Gluon exchange in a binary collision creates jets
• Jets fragment into hadrons, dominantly at
  mid-rapidity

(mini)jet
(mini)jet
26
Multiple Collisions with Nuclei

• Nuclei are extended
• RAu 6.4 fm (10-15 m)
• cf. Rp .8 fm
• Geometrical model
• Binary collisions (Ncoll)
• Participants (Npart)
• Nucleons that interact inelastically
• Spectators (2A Npart)
• pA Npart Ncoll 1
• (Npart 6 for Au)
• AA Ncoll ? Npart4/3

b
1200
Ncoll
Npart
400
b(fm)
9
0
18
27
Hard Soft
What about non-central events?
We already expect that charged particle production
can have two components
Fraction from hard processes
proton-proton multiplicity
We can tune the relative contribution by varying
the collision centrality
Is this Description unique ?
28
Parton Saturation

• Gluons recombine at a critical density
  characterized by saturation scale Qs2
• Below this scale, the nucleus looks black to a
  probe

• Gluons below x1/(2mR) overlap in transverse
  plane with size 1/Q

t
Scale depends on volume(controlled by
centrality!)
Colored Glass Condensate
McLerran, Venugopalan, Kharzeev, Dumitru,
Schaffner-Bielich
29
Data and Models for 130 GeV
Yellow band Systematic Error
30
Data and Models for 200 GeV
Yellow band Systematic Error
31
Shapes of dN/dh Distributions at 130 GeV - Hit
Counting

• Shapes only weakly dependent on centrality
• Differ in details

32
130 GeV
33
Energy Dependence and Comparison to pp

• Width increases with Ecm
• Increase D? Dybeam
• Scaling in fragmentation region

HI part. Production is increased at mid-rapidity
7-10 syst error
7-10 syst error
34
Scaling in the Fragmentation Region
UA5 Alner et al., Z. Phys. C33,1 (1986)
PHOBOS 2000/2001
7-10 syst error
Fragmentation
Fragmentation
35
Summary
Energy and Centrality Dependence of Mid-Rapidity
Multiplicity has Constrained Models and given
Insight into Interplay of Different
Processes Shapes of Multiplicity Distributions
show Scaling in Fragmentation Region illustrating
Common Mechanism for Particle Production which
Evolves to Features Unique to HI Situation at
Mid-Rapidity To Come Shapes versus Centrality
at 200 GeV Multiplicity at 20 GeV pp Data with
PHOBOS at 200 GeV