Leptonic Observables in UltraRelativistic Heavy Ion Collisions

1
Leptonic Observablesin Ultra-Relativistic Heavy
Ion Collisions
James Nagle, Columbia University for the PHENIX
Collaboration
QUARK MATTER 2002 Nantes, France July 23, 2002
2
Strangeness f ? KK- f ? ee- Dielectron
Continuum Low and Intermediate Mass
Region Charm Single Electrons (c ? D ? e
X) Quarkonia J/y ? ee- J/y ? mm-
Photons See talk of K.Reygers
3
PHENIX Experiment

• Central Arms completed in Run 2
• Detect electrons, photons, hadrons
• h lt 0.35 and pT gt 0.2 GeV/c

• South Muon Arm completed in Run 2
• Detect forward muons
• 1.2 lt h lt 2.2 and pTOT gt 2.0 GeV/c

4
PHENIX Collaboration
5
PHENIX - Run II
Au-Au running at Ecm 200 GeV RHIC delivered 42
mb-1 zlt45 cm to PHENIX We sampled with
minimum bias and Level-2 triggers 24 mb-1.
Over 50 of that in the last two weeks of the
AuAu run. Analysis show here from 26 million
minimum bias events with z lt 30 cm
Proton-Proton running at Ecm 200 GeV RHIC
delivered 700 nb-1 to PHENIX We sampled with
minimum bias and Level-1 triggers 150 nb-1.
Analysis shown here from 1.0 and 1.7 billion
proton-proton events.
6
f (1020)

• Why is the f interesting?
• Restoration of approximate chiral symmetry may
  modify the f mass and width in medium
• These modification may result in a change in the
  branching fraction of f?KK- and f?ee- when f
  decays in medium (tf 44 fm/c)
• Final state interactions of kaons may lower the
  apparent measured branching fraction of f?KK-
  relative to f?ee-.

PHENIX has excellent Particle Identification
using a high precision time-of-flight wall. Good
kaon identification makes the f?K K- measurement
possible.
See talk D.Mukhopadhyay See poster X.Li,
M.Muniruzzaman
7
f?K K-
Au Au minimum bias (0-90 central) data at
Ecm200 GeV
PHENIX preliminary
PHENIX preliminary
Yield
Yield
Mass (GeV/c2)
Mass (GeV/c2)
Signal 1135 ? 120 Signal / Background 1 /
12 Mass peak and width agree within errors of
PDG values.
8
f?K K- Results
Au Au minimum bias (0-90 central) result at
Ecm200 GeV
PHENIX Preliminary
STAR result Au Au minimum bias (0-85) at
Ecm130 GeV dN/dy 2.01 ?
0.11 (stat)
We have higher statistics using our EM
Calorimeter PID and expect mT distributions for
various centrality bins soon.
PHENIX Preliminary PID with EM Calorimeter
9
Electron Identification

• PHENIX has excellent electron identification
  capabilities.
• Ring Imaging Cherenkov Counter - threshold
  selection
• Time Expansion Chamber - dE/dx measurement
• Electromagnetic Calorimeter - Energy-Momentum
  match

All charged tracks
Apply RICH cut
Real
Net signal
Background
EMC energy / Momentum
See poster A.Lebedev
10
f ? ee-
S/B 1/20
Minimum Bias
Au Au minimum bias (0-90 central) data at
Ecm200 GeV
PHENIX preliminary
PHENIX preliminary
Yield
PHENIX preliminary
Yield
Mass (GeV/c2)
Mass (GeV/c2)
Signal / Background 1 / 20 Mass peak and width
agree within errors of PDG values.
11
f Comparison
Au Au minimum bias (0-90 central) result at
Ecm200 GeV dN/dy corrected for vacuum PDG
branching fraction values. B.F. f ? ee-
2.9 x 10-4, B.F. f?K K- 0.49
PHENIX Preliminary
PHENIX Preliminary
Data are consistent with free vacuum PDG
branching fraction values within 1s statistical
errors.
12
Dielectron Continuum

• Why is the dielectron continuum interesting?
• Low Mass Region (LMR) below the f may have excess
  dielectrons from in-medium mass modification of
  the r meson due to restoration of approximate
  chiral symmetry
• Intermediate Mass Region (IMR) above the f and
  below the J/y may have a significant contribution
  from charm at RHIC energies

Real and Mixed ee- Distribution
Real - Mixed with systematic errors
13
Low and Intermediate Mass
We compare with a calculation using free vacuum
masses and branching fractions,our measured p0,
particle ratios from proton-proton, and a charm
yield from PYTHIA scaled by the number of binary
collisions. Predictions LMR 9.2 x 10-5
IMR 1.5 x 10-5
Low Mass Region (LMR) 0.3-1.0 GeV
PHENIX Preliminary
Intermediate Mass Region (IMR) 1.1-2.5 GeV
PHENIX Preliminary
14
Single Electrons - Run I
See talk R.Averbeck
Dominant source of background electrons are
from p0 Dalitz and g conversions. We can account
for these using our published p0 measurements.
g conversion
p0 ? gee
h ? gee, 3p0
w ? ee, p0ee
Increasing excess at higher pT.
f ? ee, hee
PHENIX
r ? ee
Ratio Data / Background
h ? gee
Sys. Error
15
Subtracted Electron Spectra
PYTHIA 6.152 with CTEQ5L PDF and binary collision
scaling assumption shows good agreement with our
minimum bias and central electron distributions.
1J. Alam et al. PRC 63 (2001) 021901
16
Charm Re-Scattering
We see good agreement with our measured electrons
assuming binary scaling of PYTHIA production of
charm quarks fragmenting into D mesons that decay
into electrons.
See poster S.Batsouli
However, an alternative picture is that the charm
quarks re-scatter and take part in hydrodynamic
expansion. Using the hydro-model parameters
derived from our measured p, K, p spectra we also
find good agreement. T 121 ? 4 (stat)
MeV bS 0.70 ? 0.01(stat) We have normalized
the Hydro-charm yield using PYTHIA.
Ecm130 GeV
D from PYTHIA
D from Hydro
B from PYTHIA
B from Hydro
e from PYTHIA
e from Hydro
17
Photon Converter Method - Run II
In Run II we had a special run where we placed a
photon converter near the interaction
point. This material increases by a fixed
factor the number of electrons whose source is
photonic. Since there is a fixed relation
between g from p0 decay and p0 ?gee- Dalitz
decay, we can subtract out both
contributions. Method has many advantages and
completely independent systematic errors from
previous cocktail subtraction method.
e
?
e-
Converter
18
Run II - 200 GeV
When a large fraction of the electrons come from
non-photonic sources, the converter and
non-converter data will approach eachother. This
is observed at high pT.
19
Single Electron Results
20
Centrality Dependence
21
Observations

• Our electron data is consistent with binary
  scaling within our current statistical and
  systematic errors.
• NA50 has inferred a factor of 3 charm
  enhancement at lower energy. We do not see this
  large effect at RHIC.
• PHENIX observes a factor of 3-4 suppression in
  high pT p0 relative to binary scaling. We do not
  see this large effect in the single electrons
  from charm. Possibly less energy loss of charm
  quarks in medium due to dead-cone effect.1

NA50 - Eur. Phys. Jour. C14, 443 (2000).
Binary Scaling
PHENIX Preliminary
Enhancement of Open Charm Yield
N part
1Y.L.Dokshitzer and D.E. Kharzeev, hep-ph/0106202
22
J/y Physics

• Why is the J/y interesting?
• We expect a screening of the attractive potential
  as we approach the deconfinement transition
• This color screening may results in a decrease in
  the number of heavy quarkonia states

Lattice QCD calculation
V(r)/??
r??

• Alternative models predict enhancement from
  c-cbar coalescence
• as the collision volume cools.

See talk T.Frawley See poster H.Sato,
T.Matsumoto, X.Wei
23
J/y ? ee- in Proton-Proton
These results are from 1 billion sampled
proton-proton collisions at Ecm200 GeV using
our EM Calorimeter 2x2 tile Level-1
trigger. This represents about half our total
Run II statistics.
NJ/y 24 ? 6 (stat) ? 4 (sys)
24
Muon Measurements
North muon arm being commissioned this summer
South muon tracker (MuTR) 3 octagonal stations of
cathode strip chambers per arm
South muon identifier (MuID) 5 gaps per arm
filled with planes of transversely oriented
Iarocci tubes
See poster D.Kim, M.Liu, J.Newby, K.Read
25
J/y ? m m - in Proton-Proton
These results are from 1.7 billion sampled
proton-proton collisions at Ecm200 GeV using
our muon Level-1 trigger.
Analysis of the Au-Au data in the muon
spectrometer is underway and we expect results
soon.
26
J/y B-ds/dy (proton-proton)
27
Energy Scaling
PHENIX preliminary
CEM predictions (Phys.Lett.B390323-328,1997)
28
J/y ? ee- in Gold-Gold !
N10.8 ? 3.2 (stat) ? 3.8 (sys)
N5.9 ? 2.4 (stat) ? 0.7 (sys)
Seven different mass fitting and counting methods
used to determine systematic error in the number
of counts.
29
J/y ? ee- in Gold-Gold !
N4.5 ? 2.1 (stat) ? 0.5 (sys)
N3.5 ? 1.9 (stat) ? 0.5 (sys)
30
J/y B-dN/dy (Au-Au and p-p)
31
J/y B-dN/dy per binary collision
32
Model Comparisons
We show three different J/y patterns all
normalized to intersect our proton-proton data
point. (1) J/y scale with the number of binary
collisions (2) J/y follow normal nuclear
absorption with sJ-N7.1 mb (3) J/y follow same
pattern as NA50 (J/y / DY(mb))1
1NA50 Phys. Lett. B521, 195 (2001)
33
Hypothesis Testing Model (1)
The statistical confidence level is defined as
the probability, based on a set of measurements,
that the actual probability of an event is better
than some specified level.
34
Hypothesis Testing Model (2)
We hope to measure the nuclear absorption in d-A
at RHIC energies in the next two years. See
poster by D.Silvermyr.
35
Hypothesis Testing Model (3)
36
Conclusions
We are excited about the first results. We have
approximately a factor of two more statistics in
Au-Au from Level-2 triggers. We also hope to
have muon spectrometer results soon. We are
working on systematic errors. We have shown that
PHENIX is capable of sampling critical
physics at high rate and in a large number
of channels and are eager for a long run at RHIC
at design luminosity in Au-Au, p-p and d-A.
Just the Tip of the Iceberg