UCT Seminar VII: Scaling of Particle Yields Soft Scaling of Hard Processes

1
UCT Seminar VIIScaling of Particle
Yields(Soft Scaling of Hard Processes)

• Peter Steinberg
• Brookhaven National Laboratory
• Fulbright Scholar Program

2
Probing the Early Stages
2
Hard Probes
Saturated gluon densities
Expanding Final State
HydrodynamicFlow

• pQCD jets can be used to study medium
• Partons interact strongly with other partons
• Weakly with colorless bound hadrons

Hadron gas
QGP
3
Jet-Quenching

• While jet production can be calculated in pQCD,
  it is hard to see jets like at LEP
• Soft production fills in the gaps

• Fragmentation functions allow prediction of
  hadron spectra from pQCD cross sections
• Quenching leads to a suppression of expected
  leading hadron spectrum

4
Leading Particle Suppression
Jets modified in heavy ion collisions? -Leading
particle suppression already observed in AuAu at
130 GeV
leading particle
Partonic Energy loss in high density matter
Nuclear Modification Factor
Glauber Model
5
Why Geometry Matters

• Binary Collisions
• Jet Production
• Heavy Flavor

b
Glauber model of AA
Binary Collisions
Npart, Ncoll

• Color Exchange
• Soft Hadron Production
• Transverse Energy

Participants
wounded nucleon model
b (fm)
6
Evidence for Jet Quenching

• PHENIX results for 130 GeV
• UA1 scaled to AuAu by Ncoll
• Same spectrum as UA1 for peripheral events
• Charged particles in central events fail to scale
  with binary collisions
• Identified pions are even lower
• Nuclear shadowing Cronin effect do not modify
  result

Poster by J. Haggerty
7
Experiments!
8
The PHOBOS Apparatus

• Trigger
• Paddle Counters
• ZDCs

• 4p Multiplicity Array
• Octagon Rings
• 2-Arm Spectrometer

137000 Silicon Pad Channels
9
Whats New?
Centrality dependence of high pT hadron
suppression to pT 6 GeV/c at ?sNN 130
GeV Submitted to PRL, nucl-ex/0206011
Same analysis at ?sNN 200 GeV with comparisons
to ?sNN 130 and even higher pT
10
The device PHENIX Run- 2 EMCal

• 6 lead- scintillator (PbSc) sectors (15500
  towers)
• 2 lead- glass (PbGl) sectors (9200 towers)
• Acceptance hlt0.37 at midrapidity, f p

Pb Sc
Pb Sc
Pb Sc
Pb Sc
PbGl
Pb Gl
Pb Sc
Pb Gl
Pb Sc
PbSc
11
Results!
12
Charged Hadron Spectra
C. Jorgensen, BRAHMS Parallel Saturday
200 GeV results from all experiments
J. Jia, PHENIX Parallel Saturday
J. Klay, STAR Parallel Saturday
C. Roland, PHOBOS Parallel Saturday
13
Charged Hadron pT Spectra
J. Klay
14
Log-log Spectra
J. Klay
130 GeV nucl-ex/0206011
Power law behaviour above 3 GeV (straight line) -
pQCD
15
Comparisons with pp
16
UA1 pp Collision Spectrum
J. Klay
UA1 Measured at 200 GeV for ? lt 2.5 Reference
spectrum at 130 GeV is an interpolation of many
data sets over a wide range of ?s
Power Law Parameters 200 GeV A 31018, p0
1.770.09, n 12.080.37 130 GeV A 2676, p0
1.900.09, n 12.980.9
17
Comparing AuAu and pp Spectra
_
AuAu
How should thespectra scale?
18
Suppression of Hadron Production

• ratio of pT-spectra
• AA central / pp

K. Adcox et al. , Phys. Rev. Lett. 88, 22301
(2002)

• RAA 1 for scaling with number of binary
  collisions

J.C. Dunlop et al. , Nucl. Phys. A 698, 515c
(2002)

• RAA lt 1 for central AuAureactions at RHIC (130
  AGeV)
• observed in neutral pions and charged hadrons
  (PHENIX and STAR)

19
The results (3) pp vs AuAu (peripheral central)
p0 from PHENIX pp _at_ 200 GeV (Hisa Torii's
talk)

• Peripheral
• 
  

• Central

0-10 CENTRAL
70-80 PERIPHERAL
Ncoll 97594
Ncoll 12.3 4.0
PHENIX Preliminary
20
Jet Suppression vs. sqrt(s)
21
Theory!
22
High pT Neutral Pion Suppression Comparison To
Theory
T. Peitzmann

• pQCD calculations
• P. Levai, Nucl.Phys.A698 (2002) 631
• X.N. Wang,Phys.Rev.C61 (200) 064910
• I. Vitev, parallel talk, Monday

S. Mioduszewski, PHENIX Plenary Monday
23
Centrality Dependence
24
Comparison with NN references II

• Continues increases of suppression towards
  central collisions
• Suppression more pronounced at high pT

25
RAA Comparison to pT 6 GeV/c
J. Klay
Similar Suppression in all centralities at 200 GeV
26
Central/Peripheral Comparison
J. Klay
At 130 GeV, the suppression increases up to pT
6 GeV/c.
0.5
With higher pT data from 200 GeV, we see that the
suppression has saturated at pT 6 GeV/c
0.5
27
Centrality Dependence of Suppression
T. Peitzmann

• RAA for neutral pions as a function of centrality
• gradual decrease
• stronger decrease for higher pT
• no threshold effect
• surface-to-volume?

D. dEnterria, PHENIX Parallel Saturday
28
Participant Scaling
29
Another way to compare to pp?
_
30
Number of collisions at different Energies
31
Scaled Spectra / pp-Fit (Npart!)
_
PHOBOS Preliminary

• Centrality range
• ltbgt from 10 to 3 fm
• ltngt from 3 to 6

32
Participant vs. Collision Scaling?
T. Peitzmann
C. Roland, PHOBOS Parallel Saturday
J. Jia, PHENIX Parallel Saturday
33
Neutral Pion Suppression as Participant Scaling?
T. Peitzmann

• participant scaling valid in high pT limit?
• participant scaling works better than
  collision scaling
• accidental?
• surface emission?
• does not describe scaling from pp to AA!
• not claimed by PHOBOS

Comparison of central to (semi-)peripheral is
instructive, but Central AA may be seen as
superposition of NN collisions (in independent
collision picture), not as superposition of
peripheral AA
Whos afraid of participant scaling? Relax !
Interesting observation (not trivial at high
pT), but not necessarily physics!
34
Spectral Modification vs. Npart
35
Charged particle pT spectra from 200 GeV
pT gt2 GeV/c, decrease of inverse slope ?
suppression
h h-
pT lt2 GeV/c, increase of inverse slope ? flow

• In bins of centrality, spectra show a gradual
  loss in concavity as they evolve from peripheral
  to central collisions.
• pTlt2 GeV/c, increase of inverse slope
• pTgt2 GeV/c, inverse slope decrease

Collective flow
Jet quenched
36
Minbias Ratios at ?sNN 200 GeV
Why do it? Minimizes the uncertainty due to the
reference spectrum
Minimum bias is most like 30-40
Broader pT range
37
Evolution with Centrality
PHOBOS

• Follow change of shape vs most peripheral bin.

Modulation aroundNpart scaling?
38
Conclusions

• Jet quenching predicted and found!
• High pT does not scale like pp x Ncoll
• Not much variation with pT
• Participant scaling is an efficient description
  of the existing data
• However, no good physical picture
• Tough to get high-pT from soft string
  fragmentation
• Existing theory is having trouble with this data

39
The End
40
Charged Particle Production
Central Density
200 GeV
AuAu
n Ncoll/(Npart/2)
130 GeV
19.6 GeV preliminary
Data from PRC 65 061901R (2002)
Two Component Model
41
Charged Particle Production
Total Multiplicity
Preliminary
from Peter Steinbergs talk

• Total multiplicity scales with Npart

42
How to compare spectra to pp?
_

• Observations
• Mid-rapidity multiplicity compatible with two
  component model
• Total multiplicity shows Npart scaling
• How do spectra scale with centrality?
• Does scaling change over the pT range?
• Crucial for physics interpretation

43
Scaled Spectra / pp-Fit
_
PHOBOS Preliminary

• Shape differs from pp already at Npart 65
• Moderate change from Npart 65 to Npart 344

_
44
Centrality scaling in pT bins
Relative Yield
Npart
Spectra normalized to yield at Npart 65
45
Npart Scaling at high pT
PHOBOS Preliminary
Ncoll-scaling
Normalized to yield at Npart 65

• Npart scaling describes data at pT 4.25 GeV/c

46
Summary

• Measured h,- pT spectra in 200 GeV AuAu
  collisions
• Transverse momentum 0.2 lt pT lt 5 GeV/c
• Centrality range ltNpartgt from 65 to 344
• Data show
• Substantial difference in spectral shape between
  pp and peripheral AuAu (Npart 65)
• Minor change from 65 to 344 participants
• Even at pT of 4 5 GeV/c, Npart -scaling from
  peripheral to central AuAu

_
47
Comparison to Lower Energies

• Data taken at 130GeV shows similar trends
• Shape is consistent with measurements by STAR

Relative Yield
48
Integrated Yields vs Centrality
dN/dy Spectra Integral
dN/dh - Multiplicity data
49
Results (5) Ratios Central/Peripheral
(part of the uncertainties cancel out)

• Ratio AuAu(different centralities) / AuAu(70-80)
  

(yellow bands uncertainties in binary scaling)
PHENIX Preliminary
Clear suppression observed in central collisions
compared to periph.
50
Results (6) Nuclear modification factor

• Ratio AuAu / (Ncoll scaled) pp

pp
Cronin enhancement
0-10 Central
(Ncoll97594)
PHENIX Preliminary
30-40 Semi-central
RAA(2 GeV/c) 0.45
(Ncoll22014)
RAA(8 GeV/c) 0.16
Behaviour at high pT (x2 pT/vs 0.1)
inconsistent with initial- state nuclear effects
(shadowing ? antishadowing).
51
Results (7) Onset of suppression ?

• RAA plotted as a function of centrality
  transverse energy(1) ( eBjorken)
• Suppression sets in for
• centrality 40-60

Transverse energy measured in EMCal as a
function of centrality (S.Bazilevsky talk)
(t0 1 fm/c) pR2 Centrality
40- 50
PHENIX Preliminary
30- 40
20- 30
10- 20
0- 10
PHENIX Preliminary
(1) ET See S. Bazilevsky talk
0- 10
Bjorken Energy density
52
Charged Hadron pT Spectra
pp at 200 GeV Charged hadron Spectrum measured
by STAR to high pT However Absolute
normalization and trigger bias corrections being
finalized STAR Reference spectrum in progress
53
?sNN Spectra Ratio
Ratio of 200 GeV Spectra to 130 GeV Spectra
For comparison, UA1(200)/Extrap(130) for ? lt 0.5
Most peripheral centrality determination
systematics still need to be understood
54
NN Reference Discussion
Essential Ingredients
130 GeV
UA1 Acceptance ? ? lt 2.5 STAR Acceptance ? ?
lt 0.5 Non-trivial! Must be taken into
account. Reminder UA1 spectra are from pp
collisions, but isospin corrections are small ?
negligible few
dNch/d? (?lt0.5) / (? lt 2.5)
200 GeV
pT (GeV/c)
55
The PHOBOS Spectrometer
10cm

• Outer layers situated in 2T magnetic field
• High segmentation in bending direction
• Tracking within 10 cm of interaction point
• Coverage near mid-rapidity
• Phi acceptance of 3 per Arm

70 cm
z
x
y
56
Particle Tracking In Spectrometer

• Road-following algorithm finds straight tracks in
  field-free region
• Curved tracks in B-field found by clusters in
  (1/p, ?) space
• Match pieces by ?, consistency in dE/dx and fit
  in yz-plane
• Covariance Matrix Track Fit for momentum
  reconstruction and ghost rejection

z
By
Beam
2
1
x
10 cm
57
Spectrometer Performance
Acceptance
Momentum Resolution

• Data Sample Production Run 2001(200 GeV)
• 7.8 M AuAu Events, Min. Bias Trigger
• 32 M reconstructed particles

58
Efficiency Acceptance
59
AuAu Analysis
High pT Track Selection
Event Selection
dca ? distance of closest approach to the primary
vertex
Primary vertex
Primary vertex
signed Dca lt 1 cm
Centrality classes based on mid-rapidity
multiplicity
? lt 0.5
60
Momentum Resolution Corrections
0.5 Tesla Field factor of 3 improvement in pT
resolution compared to 0.25 Tesla field (130 GeV
data)
(Trivial field factor Smaller diffusion ?
smaller hits, smaller residuals, better
determined momenta)
61
Background Contamination

• How do we get rid of whats left?
• Hijing, scaled to match STAR measured data at
  130 GeV (1.14yield130Gev and 1.04slope130GeV
  for 200 GeV), extrapolated to high pT

Most significant contribution to systematic
uncertainty at high pT
Above 3 GeV, the corrections are small 10, but
we apply 100 error due to extrapolation
uncertainty
62
The probe p0

• Energy scale and resolution gg
  invariant-mass analysis
• 
  p0 peak at
  pT gt 8 GeV/c
• 
  (min.bias)
• (Excellent agreement real data
• and embedded single p0 )

GeV
GeV
Counts
63
The systematics (1) Correction factors syst.
errors

• Efficiency losses calculated with 2M simulated
  single p0 embedded in real data
• Correction factors x20.- x10.
• Acceptance 1/0.25
• Efficiency 1/0.20- 0.30
• Systematic errors ( yield)
• p 0 extraction 15
• pT smearing 10
• TOF cut 10
• Fiduc., asym., hot towers cuts 5
• Acceptance 3
• Off-vertex p0 contribution 3
• Final systematic error 20- 30 (periph),
  23- 33 (central) increasing with pT

Acceptance x Efficiency Correction Factors